<p>Artificial Intelligence firm Anthropic recently showed some US lawmakers a demo of its Mythos model, which could identify vulnerabilities in the websites, intranets, and online systems of banking and financial institutions, among other things. The model could either fix those vulnerabilities or exploit them, potentially enabling the theft of funds from bank accounts. Andrew Garbarino, one of the lawmakers who attended the demonstration, has now spoken publicly, expressing alarm at what was shown.</p><h2>What House Homeland Security Committee Chair Andrew Garbarino said about Anthropic's controversial Mythos model</h2><p>"They showed us what it could do, and it's scary," Representative Andrew Garbarino, a New York Republican and Chair of the House Homeland Security Committee, told Punchbowl News during its "Fly Out Day" programme.</p><p>Garbarino said the closed-door demonstration left him shaken. During the demo, Anthropic "told the model to find a vulnerability in a bank and empty accounts, and then it went and did it," Garbarino told Punchbowl. He added that Mythos was then able to identify the vulnerability and fix it.</p><h2>Republican Congressman Garbarino wants the government to act on AI</h2><p>Garbarino said he wants Washington to move quickly on AI. The Anthropic demo showed that the government must "act soon", he said, while also acknowledging the challenges involved.</p><p>"I say act, but I don't know what the answer is," Garbarino said, admitting that lawmakers have been slow to respond to developments in AI. "We don't know what we have to do," he said, adding, "I'd say 95 per cent of my colleagues don't understand what the hell's going on."</p><p><a href="https://www.wionews.com/videos/us-treasury-to-test-anthropic-s-mythos-for-cyber-risks-1776229191378/embed">https://www.wionews.com/videos/us-treasury-to-test-anthropic-s-mythos-for-cyber-risks-1776229191378/embed</a></p><h2>Momentum builds towards AI regulation</h2><p>Garbarino praised US President Donald Trump's executive order directing federal agencies to prepare for AI-driven cyber threats. He also backed a discussion draft released by Representatives Jay Obernolte and Lori Trahan proposing a federal AI regulatory framework. According to Garbarino, government agencies and Congress should be given access to such advanced AI platforms before they are released to the public.</p><p>Also read: <a href="https://www.wionews.com/technology/anthropic-s-most-powerful-ai-model-gets-partial-us-approval-here-s-what-changed-1782540188737">Anthropic's most powerful AI model gets partial US approval. Here's what changed</a></p><h2>Demos by Anthropic, OpenAI and others are ringing alarm bells</h2><p>Details remain limited, but the session described by Garbarino was a private demonstration rather than a public hearing. It formed part of a broader series of briefings for US lawmakers and officials. Axios reported in April that Anthropic and OpenAI had briefed House Homeland Security Committee staff on their cyber-capable AI models during classified sessions.</p><p>Garbarino's committee has also held public hearings on AI-related security issues, including a December 2025 hearing prompted by Anthropic's disclosure of a largely autonomous cyberattack linked to China.</p><p>Anthropic officials also demonstrated a so-called "jailbreak", in which an AI model bypasses the safety guardrails imposed by its human developers. According to Garbarino, one unspecified model generated a plan to kidnap a lawmaker in just 30 seconds. He warned that malicious actors could use similar tools to "turn off this gas pipeline" or "increase the amount of chlorine" in a public water system. "You have the ability to not just learn how to do it, but to tell it to do it for you," he said.</p><p>Also read: <a href="https://www.wionews.com/world/after-anthropic-us-now-limits-openai-s-latest-powerful-ai-model-to-domestic-users-1782514231980https://www.wionews.com/world/after-anthropic-us-now-limits-openai-s-latest-powerful-ai-model-to-domestic-users-1782514231980">After Anthropic, US now limits OpenAI's latest powerful AI model to domestic users</a></p><h2>What is Mythos, and how dangerous is it?</h2><p>Mythos is reported to be Anthropic's most advanced AI model to date. Owing to its capabilities, the company has reportedly withheld it from broad public release. The concerns centres on its ability to identify and exploit security flaws in websites, enterprise networks and other online systems. According to reports, the White House has restricted public access to Mythos and directed Anthropic to limit access to some of its most advanced AI models, including by restricting availability to certain foreign nationals, due to national security concerns.</p><p>The Associated Press news agency recently reported that Mythos was able to identify vulnerabilities in classified US government systems as part of Anthropic's Project Glasswing, a limited deployment of the model to selected organisations for testing. During a US Senate hearing on 11 June, Senator Mark Warner said the tool "broke into almost all of our classified systems, not in weeks but in hours", underscoring lawmakers' concerns about the cybersecurity risks posed by highly capable AI systems.</p>

US Congressman Andrew Garbarino revealed in an interview that Anthropic's Mythos AI model demonstrated the ability to find and exploit banking vulnerabilities. He is asking for regulation. So are many others in the government and private sectors

Artificial Intelligence firm Anthropic recently showed some US lawmakers a demo of its Mythos model, which could identify vulnerabilities in the websites, intranets, and online systems of banking and financial institutions, among other things. The model could either fix those vulnerabilities or exploit them, potentially enabling the theft of funds from bank accounts. Andrew Garbarino, one of the lawmakers who attended the demonstration, has now spoken publicly, expressing alarm at what was shown. What House Homeland Security Committee Chair Andrew Garbarino said about Anthropic's controversial Mythos model "They showed us what it could do, and it's scary," Representative Andrew Garbarino, a New York Republican and Chair of the House Homeland Security Committee, told Punchbowl News during its "Fly Out Day" programme. Garbarino said the closed-door demonstration left him shaken. During the demo, Anthropic "told the model to find a vulnerability in a bank and empty accounts, and then it went and did it," Garbarino told Punchbowl. He added that Mythos was then able to identify the vulnerability and fix it. Republican Congressman Garbarino wants the government to act on AI Garbarino said he wants Washington to move quickly on AI. The Anthropic demo showed that the government must "act soon", he said, while also acknowledging the challenges involved. "I say act, but I don't know what the answer is," Garbarino said, admitting that lawmakers have been slow to respond to developments in AI. "We don't know what we have to do," he said, adding, "I'd say 95 per cent of my colleagues don't understand what the hell's going on." https://www.wionews.com/videos/us-treasury-to-test-anthropic-s-mythos-for-cyber-risks-1776229191378/embed Momentum builds towards AI regulation Garbarino praised US President Donald Trump's executive order directing federal agencies to prepare for AI-driven cyber threats. He also backed a discussion draft released by Representatives Jay Obernolte and Lori Trahan proposing a federal AI regulatory framework. According to Garbarino, government agencies and Congress should be given access to such advanced AI platforms before they are released to the public. Also read: Anthropic's most powerful AI model gets partial US approval. Here's what changed Demos by Anthropic, OpenAI and others are ringing alarm bells Details remain limited, but the session described by Garbarino was a private demonstration rather than a public hearing. It formed part of a broader series of briefings for US lawmakers and officials. Axios reported in April that Anthropic and OpenAI had briefed House Homeland Security Committee staff on their cyber-capable AI models during classified sessions. Garbarino's committee has also held public hearings on AI-related security issues, including a December 2025 hearing prompted by Anthropic's disclosure of a largely autonomous cyberattack linked to China. Anthropic officials also demonstrated a so-called "jailbreak", in which an AI model bypasses the safety guardrails imposed by its human developers. According to Garbarino, one unspecified model generated a plan to kidnap a lawmaker in just 30 seconds. He warned that malicious actors could use similar tools to "turn off this gas pipeline" or "increase the amount of chlorine" in a public water system. "You have the ability to not just learn how to do it, but to tell it to do it for you," he said. Also read: After Anthropic, US now limits OpenAI's latest powerful AI model to domestic users What is Mythos, and how dangerous is it? Mythos is reported to be Anthropic's most advanced AI model to date. Owing to its capabilities, the company has reportedly withheld it from broad public release. The concerns centres on its ability to identify and exploit security flaws in websites, enterprise networks and other online systems. According to reports, the White House has restricted public access to Mythos and directed Anthropic to limit access to some of its most advanced AI models, including by restricting availability to certain foreign nationals, due to national security concerns. The Associated Press news agency recently reported that Mythos was able to identify vulnerabilities in classified US government systems as part of Anthropic's Project Glasswing, a limited deployment of the model to selected organisations for testing. During a US Senate hearing on 11 June, Senator Mark Warner said the tool "broke into almost all of our classified systems, not in weeks but in hours", underscoring lawmakers' concerns about the cybersecurity risks posed by highly capable AI systems.

<p>US Congressman Andrew Garbarino revealed that Anthropic's Mythos AI model demonstrated the ability to find and exploit banking vulnerabilities. He is asking for regulation. So are many others in the government and private sectors</p>

'They showed us what it could do, and it's scary': Congressman on Anthropic Mythos model demo

In a major policy shift, the U.S. government has authorized the redeployment of Anthropic’s advanced AI model, Mythos 5, to more than 100 vetted organizations. This decision marks a significant de-escalation of the national security standoff that led to a total suspension of the model just two weeks ago.


US clears mythos 5 for 100+ trusted organization's

<p>"Did India miss the AI bus?" This is a frequently asked question at recent artificial intelligence-related events in India. Clearly, the progress has been slow. But what is the reason? The answer may lie in how low, and slow, the investment has been in AI startups, foundational AI model development, and research and development (R&amp;D). Here is a status report:</p><h2>India's research and development funding gap</h2><p>India's gross expenditure on research and development (GERD) is around 0.64 per cent of GDP. This is below the global average and far behind China, at around 2.6 per cent, and the US, at around 3.6 per cent, according to the <a href="https://power.lowyinstitute.org/data/economic-capability/technology/rnd-spending-of-gdp/">Lowy Institute</a>. Private industry funds only about 36 per cent of total R&amp;D in India, compared with more than 70 per cent in the US, China and South Korea, according to a Carnegie Endowment for International Peace analysis. It also noted that India trails the two top AI leader nations in AI patents granted, reflecting not only a spending gap but also a quality gap.</p><p>Also read: <a href="https://www.wionews.com/india-news/india-s-digital-paradox-world-class-talent-third-world-capital-wion-expert-speak-1781084191225">India's digital paradox: World-class talent, third-world capital | WION Expert Speak</a></p><h2>Investment in AI startups in India reveals a capital chasm</h2><p>Indian AI startups raised $643 million across 100 deals in 2025, up just 4.1 per cent, while AI funding in US startups surged 141 per cent to $121 billion across 765 rounds, according to Tracxn data reported by <a href="https://techcrunch.com/2025/12/27/india-startup-funding-hits-11b-in-2025-as-investors-grow-more-selective/">TechCrunch</a>. Indian capital largely favoured application-led firms over capital-intensive foundational model developers. Investors attribute this to India's lack of large foundational AI model companies and the time required to build research depth, talent pipelines and patient capital.</p><p>However, there has been a sharp upturn in 2026. Indian AI startups raised <a href="https://www.analyticsinsight.net/amp/story/startups/how-indian-ai-startups-raised-394-billion-in-just-three-months">$3.94 billion</a> in the first quarter of this year alone, more than six times the full-year 2025 total. This was led by Neysa's $600 million funding round and the momentum generated by the AI Impact Summit hosted by India, according to Analytics Insight.</p><p>New Delhi has reinforced this shift with a $1.15 billion Fund of Funds and a Rs 1 trillion ($12 billion) Research, Development and Innovation scheme spanning AI, quantum computing and deep-tech technologies, as reported by TechCrunch.</p><p>Much of this remains at the policy announcement and reporting stage, so its long-term durability remains to be seen. However, it is fair to say that the direction has clearly reversed.</p><p>Also read: <a href="https://www.wionews.com/india-news/india-s-ai-problem-gaps-in-compute-hardware-infrastructure-chips-1781943963441">India's AI problem: Gaps in compute, hardware, infrastructure, chips affect readiness</a></p><h2>Research output: China leads, India climbs slowly</h2><p>China dominates high-impact research, with the Australian Strategic Policy Institute's (<a href="https://www.aspistrategist.org.au/aspis-critical-technology-tracker-2025-updates-and-10-new-technologies/">ASPI</a>) Critical Technology Tracker, published in December 2025, showing China leading in 66 of 74 critical technologies like AI. India is slowly climbing up, having ranked 50th, up from 43rd a year earlier. India's weakness is concentrated in compute infrastructure and frontier AI model categories that matter most, rather than in the breadth of its research base.</p><p><a href="https://www.wionews.com/videos/how-india-is-positioning-itself-in-an-ai-world-wion-1771597788504/embed">https://www.wionews.com/videos/how-india-is-positioning-itself-in-an-ai-world-wion-1771597788504/embed</a></p><h2>Foundational AI: India's sovereign model bet</h2><p>India's foundational AI model push has been stuck in the planning stage for years, but is now moving towards <a href=" https://explainx.ai/blog/india-sovereign-ai-status-indiaai-mission-2026">execution</a>. <a href=" https://www.sarvam.ai/blogs/sarvam-30b-105b">Sarvam AI </a>was selected under the IndiaAI Mission in April 2025 to build the country's first sovereign large language model (LLM) ecosystem. Sarvam's original mandate was to build a 120-billion-parameter model. It <a href="https://www.opensourceforu.com/2026/03/sarvam-releases-30b-and-105b-llms-under-apache-2-0/">released</a> two open-source models in early 2026, Sarvam-30B and Sarvam-105B, under the Apache 2.0 licence on Hugging Face and AIKosh. Both are mixture-of-experts (MoE) reasoning models trained from scratch using government-supported compute infrastructure. A separate awardee, Soket AI, retains the original 120-billion-parameter model mandate.</p><p>Also read: <a href="https://www.wionews.com/technology/forget-building-ai-india-is-yet-to-build-its-first-fully-indigenous-computer-or-mobile-phone-1781750813969">Forget building AI, India is yet to build its first fully indigenous computer or mobile phone</a></p><p><br>These should be viewed alongside the limitations. Sarvam's models were trained on a few thousand graphics processing units (GPUs) over several months, far below the cluster-months of more than 10,000 GPUs typically required for frontier-scale AI training. In addition, all sovereign model training continues to rely on imported NVIDIA chips, highlighting India's dependence on foreign semiconductor hardware. While India has now demonstrated that it can build and release competitive open-source AI models, closing the capability gap in frontier-scale AI systems remains the more difficult challenge.</p>

India's AI progress has been constrained by low R&D spending and limited startup funding. The IndiaAI Mission and sovereign AI model efforts signal a significant shift in direction. Here is a status report

"Did India miss the AI bus?" This is a frequently asked question at recent artificial intelligence-related events in India. Clearly, the progress has been slow. But what is the reason? The answer may lie in how low, and slow, the investment has been in AI startups, foundational AI model development, and research and development (R&D). Here is a status report: India's research and development funding gap India's gross expenditure on research and development (GERD) is around 0.64 per cent of GDP. This is below the global average and far behind China, at around 2.6 per cent, and the US, at around 3.6 per cent, according to the Lowy Institute. Private industry funds only about 36 per cent of total R&D in India, compared with more than 70 per cent in the US, China and South Korea, according to a Carnegie Endowment for International Peace analysis. It also noted that India trails the two top AI leader nations in AI patents granted, reflecting not only a spending gap but also a quality gap. Also read: India's digital paradox: World-class talent, third-world capital | WION Expert Speak Investment in AI startups in India reveals a capital chasm Indian AI startups raised $643 million across 100 deals in 2025, up just 4.1 per cent, while AI funding in US startups surged 141 per cent to $121 billion across 765 rounds, according to Tracxn data reported by TechCrunch. Indian capital largely favoured application-led firms over capital-intensive foundational model developers. Investors attribute this to India's lack of large foundational AI model companies and the time required to build research depth, talent pipelines and patient capital. However, there has been a sharp upturn in 2026. Indian AI startups raised $3.94 billion in the first quarter of this year alone, more than six times the full-year 2025 total. This was led by Neysa's $600 million funding round and the momentum generated by the AI Impact Summit hosted by India, according to Analytics Insight. New Delhi has reinforced this shift with a $1.15 billion Fund of Funds and a Rs 1 trillion ($12 billion) Research, Development and Innovation scheme spanning AI, quantum computing and deep-tech technologies, as reported by TechCrunch. Much of this remains at the policy announcement and reporting stage, so its long-term durability remains to be seen. However, it is fair to say that the direction has clearly reversed. Also read: India's AI problem: Gaps in compute, hardware, infrastructure, chips affect readiness Research output: China leads, India climbs slowly China dominates high-impact research, with the Australian Strategic Policy Institute's (ASPI) Critical Technology Tracker, published in December 2025, showing China leading in 66 of 74 critical technologies like AI. India is slowly climbing up, having ranked 50th, up from 43rd a year earlier. India's weakness is concentrated in compute infrastructure and frontier AI model categories that matter most, rather than in the breadth of its research base. https://www.wionews.com/videos/how-india-is-positioning-itself-in-an-ai-world-wion-1771597788504/embed Foundational AI: India's sovereign model bet India's foundational AI model push has been stuck in the planning stage for years, but is now moving towards execution. Sarvam AI was selected under the IndiaAI Mission in April 2025 to build the country's first sovereign large language model (LLM) ecosystem. Sarvam's original mandate was to build a 120-billion-parameter model. It released two open-source models in early 2026, Sarvam-30B and Sarvam-105B, under the Apache 2.0 licence on Hugging Face and AIKosh. Both are mixture-of-experts (MoE) reasoning models trained from scratch using government-supported compute infrastructure. A separate awardee, Soket AI, retains the original 120-billion-parameter model mandate. Also read: Forget building AI, India is yet to build its first fully indigenous computer or mobile phone These should be viewed alongside the limitations. Sarvam's models were trained on a few thousand graphics processing units (GPUs) over several months, far below the cluster-months of more than 10,000 GPUs typically required for frontier-scale AI training. In addition, all sovereign model training continues to rely on imported NVIDIA chips, highlighting India's dependence on foreign semiconductor hardware. While India has now demonstrated that it can build and release competitive open-source AI models, closing the capability gap in frontier-scale AI systems remains the more difficult challenge.

<p>India's AI progress has been constrained by low R&amp;D spending and limited startup funding. The IndiaAI Mission and sovereign AI model efforts signal a significant shift in direction. Here is a status report</p>

India's AI problem: Funding gap for AI startups, research and development is the core constraint

The global AI boom is reshaping more than just the technology industry; it's changing South Korea's social hierarchy. Employees at chipmaking giants Samsung Electronics and SK Hynix are now among the country's most sought-after professionals, thanks to soaring salaries and performance bonuses driven by record demand for AI memory chips.


The AI boom turns chip workers into South Korea's new elite

<p>What could be called 'artificial intelligence cannibalism' is allegedly being committed by Chinese AI firms targeting American AI company Anthropic. In a recent formal complaint, Anthropic publicly accused Chinese tech giant Alibaba's Qwen AI lab of conducting the largest known "adversarial distillation attack" to extract capabilities from its widely used Claude AI models. How does AI distillation work? Here is what you should know.</p><h2>What is Anthropic accusing Alibaba of doing?</h2><p>Ahead of the upcoming US Senate hearing on artificial intelligence, Anthropic sent a letter on &nbsp;June 10 to US Senators Tim Scott and Elizabeth Warren of the Senate Committee on Banking, Housing, and Urban Affairs, <a href="https://www.reuters.com/world/china/anthropic-says-alibaba-illicitly-extracted-claude-ai-model-capabilities-2026-06-24/">alleging </a>that millions of ‘AI distillation’ attacks were carried out by Alibaba's AI wing on Claude.</p><p>The letter claimed that operators linked to Alibaba and its Qwen AI lab used approximately 25,000 fraudulent accounts to generate more than 28.8 million exchanges with Claude models. These exchanges consisted of interactions and prompts submitted to the AI model over a six-week period between 22 April and 5 June 2026. &nbsp;Anthropic described the attacks as an effort to "brazenly and illicitly" extract Claude's capabilities and called it the <a href="https://www.wsj.com/tech/ai/anthropic-claims-alibaba-ran-brazen-campaign-to-access-its-claude-ai-model-69d7a392">largest known </a>distillation attack against the company to date.</p><p>Also read: <a href="https://www.wionews.com/technology/are-ai-tools-reducing-people-s-critical-thinking-skills-1777272249863">Are AI tools reducing people’s critical thinking skills?</a></p><h2>What is an 'adversarial distillation attack'?</h2><p>An adversarial distillation attack is a technique in which a stronger AI model—in this case Claude—is queried at scale to collect high-quality responses and outputs from it. These responses are then used to train or improve weaker or less capable AI models. This method allows attackers to replicate advanced capabilities of large language models (LLMs), including reasoning, coding and agentic behaviour, at a fraction of the original training cost, without direct access to the model's weights or architecture.</p><p>AI model distillation is increasingly being used as a low-cost method by competitors, particularly those facing computing or AI chip-related sanctions, to narrow the gap with leading US AI laboratories such as Anthropic, OpenAI and Google. Anthropic has been increasingly vocal about these threats.</p><p>Also read: <a href="https://www.wionews.com/technology/ai-is-making-your-next-macbook-more-expensive-here-s-why-1782458264297">AI is making your next MacBook more expensive. Here's why</a></p><h2>Alibaba allegedly targeted Anthropic's agentic reasoning, autonomous and coding capabilities</h2><p>In the letter, Anthropic claimed that the alleged distillation campaign from China specifically targeted Claude's most valuable and commercially sensitive capabilities, including software engineering, coding, agentic reasoning, autonomous decision-making and task handling. According to the company, Claude's ability to perform long-horizon and complex tasks, as well as its advanced reasoning capabilities, were also targeted.</p><p>Anthropic said these actions violated its terms of service and geographic access restrictions, given that Claude is not officially available in China, while allegedly ignoring repeated US government warnings about AI intellectual property theft.</p><p><a href="https://www.wionews.com/videos/debate-continues-over-ai-s-impact-on-human-jobs-summer-davos-1782373302795/embed">https://www.wionews.com/videos/debate-continues-over-ai-s-impact-on-human-jobs-summer-davos-1782373302795/embed</a></p><h2>Past instances of AI distillation attacks targeting Anthropic</h2><p>This is not the first time that American AI companies have faced alleged distillation attacks. In February 2026, Anthropic publicly identified similar, although smaller-scale, campaigns by other Chinese AI labs. The company allegedly recorded around 150,000 exchanges from DeepSeek, more than 3.4 million exchanges from Moonshot AI, and approximately 13 million exchanges from MiniMax. These campaigns were significantly smaller than Alibaba's alleged 28.8 million exchanges.</p><p>Also read: <a href="https://www.wionews.com/world/ai-washing-is-here-fake-product-claims-citing-ai-as-cause-of-layoffs-and-more-and-no-ai-won-t-cure-diseases-doctors-will-1782370454223">'AI Washing': Fake product claims, citing AI as cause of layoffs, and more. And no, AI won't cure diseases, doctors will</a></p><h2>White House accusations of Chinese "industrial-scale" AI intellectual property theft</h2><p>Shortly after Anthropic sent the letter, the US Department of Commerce, on June 12, imposed restrictions on access to the company's newest AI models, Mythos and Fable, citing concerns that they could be used by military and intelligence entities in China and other countries of concern. Anthropic subsequently disabled global access to those models.</p><p>Around the same time, Alibaba was added to the US Department of Defense's list of Chinese military companies, alongside several other Chinese firms. Alibaba is challenging that designation.</p><p>Anthropic has called for coordinated government-industry action, greater threat intelligence sharing and stronger penalties to protect US leadership in artificial intelligence.</p><p>The company has also expressed support for US government's efforts to combat AI intellectual property theft and reaffirmed its commitment to working with Congress and the administration on the issue. Alibaba's shares reportedly fell by around 3 per cent following news of the allegations, while the controversy renewed debate over national security, intellectual property protection and the global AI arms race.</p>

Anthropic accused Alibaba's Qwen AI lab of carrying out the largest known AI 'distillation attack' on Claude. Millions of interactions were used to extract advanced AI capabilities, it claimed. But what is a distillation attack?

What could be called 'artificial intelligence cannibalism' is allegedly being committed by Chinese AI firms targeting American AI company Anthropic. In a recent formal complaint, Anthropic publicly accused Chinese tech giant Alibaba's Qwen AI lab of conducting the largest known "adversarial distillation attack" to extract capabilities from its widely used Claude AI models. How does AI distillation work? Here is what you should know. What is Anthropic accusing Alibaba of doing? Ahead of the upcoming US Senate hearing on artificial intelligence, Anthropic sent a letter on June 10 to US Senators Tim Scott and Elizabeth Warren of the Senate Committee on Banking, Housing, and Urban Affairs, alleging that millions of ‘AI distillation’ attacks were carried out by Alibaba's AI wing on Claude. The letter claimed that operators linked to Alibaba and its Qwen AI lab used approximately 25,000 fraudulent accounts to generate more than 28.8 million exchanges with Claude models. These exchanges consisted of interactions and prompts submitted to the AI model over a six-week period between 22 April and 5 June 2026. Anthropic described the attacks as an effort to "brazenly and illicitly" extract Claude's capabilities and called it the largest known distillation attack against the company to date. Also read: Are AI tools reducing people’s critical thinking skills? What is an 'adversarial distillation attack'? An adversarial distillation attack is a technique in which a stronger AI model—in this case Claude—is queried at scale to collect high-quality responses and outputs from it. These responses are then used to train or improve weaker or less capable AI models. This method allows attackers to replicate advanced capabilities of large language models (LLMs), including reasoning, coding and agentic behaviour, at a fraction of the original training cost, without direct access to the model's weights or architecture. AI model distillation is increasingly being used as a low-cost method by competitors, particularly those facing computing or AI chip-related sanctions, to narrow the gap with leading US AI laboratories such as Anthropic, OpenAI and Google. Anthropic has been increasingly vocal about these threats. Also read: AI is making your next MacBook more expensive. Here's why Alibaba allegedly targeted Anthropic's agentic reasoning, autonomous and coding capabilities In the letter, Anthropic claimed that the alleged distillation campaign from China specifically targeted Claude's most valuable and commercially sensitive capabilities, including software engineering, coding, agentic reasoning, autonomous decision-making and task handling. According to the company, Claude's ability to perform long-horizon and complex tasks, as well as its advanced reasoning capabilities, were also targeted. Anthropic said these actions violated its terms of service and geographic access restrictions, given that Claude is not officially available in China, while allegedly ignoring repeated US government warnings about AI intellectual property theft. https://www.wionews.com/videos/debate-continues-over-ai-s-impact-on-human-jobs-summer-davos-1782373302795/embed Past instances of AI distillation attacks targeting Anthropic This is not the first time that American AI companies have faced alleged distillation attacks. In February 2026, Anthropic publicly identified similar, although smaller-scale, campaigns by other Chinese AI labs. The company allegedly recorded around 150,000 exchanges from DeepSeek, more than 3.4 million exchanges from Moonshot AI, and approximately 13 million exchanges from MiniMax. These campaigns were significantly smaller than Alibaba's alleged 28.8 million exchanges. Also read: 'AI Washing': Fake product claims, citing AI as cause of layoffs, and more. And no, AI won't cure diseases, doctors will White House accusations of Chinese "industrial-scale" AI intellectual property theft Shortly after Anthropic sent the letter, the US Department of Commerce, on June 12, imposed restrictions on access to the company's newest AI models, Mythos and Fable, citing concerns that they could be used by military and intelligence entities in China and other countries of concern. Anthropic subsequently disabled global access to those models. Around the same time, Alibaba was added to the US Department of Defense's list of Chinese military companies, alongside several other Chinese firms. Alibaba is challenging that designation. Anthropic has called for coordinated government-industry action, greater threat intelligence sharing and stronger penalties to protect US leadership in artificial intelligence. The company has also expressed support for US government's efforts to combat AI intellectual property theft and reaffirmed its commitment to working with Congress and the administration on the issue. Alibaba's shares reportedly fell by around 3 per cent following news of the allegations, while the controversy renewed debate over national security, intellectual property protection and the global AI arms race.

<p>Anthropic accused Alibaba's Qwen AI lab of carrying out the largest known AI 'distillation attack' on its AI Claude model. Millions of interactions were used to extract advanced AI capabilities, it claimed. But what is a distillation attack?</p>

‘AI Cannibalism’: Anthropic accuses Chinese tech giant Alibaba of 29 million adversarial distillation attacks. What is it?

<p>&nbsp;</p><p>Artificial intelligence is making computers smarter. It is also making them more expensive. Apple recently increased prices for several MacBook and iPad models after memory costs surged, highlighting a growing problem across the technology industry. The company says demand for memory chips has risen so quickly that it has become difficult for consumer electronics makers to secure enough supply at reasonable prices. The shortage is largely being driven by AI companies building massive data centres to train and run increasingly powerful artificial intelligence models.</p><h2><strong>Why AI needs so much memory</strong></h2><p>Every AI model relies on enormous amounts of high-speed memory. Whether it is ChatGPT, Gemini, Claude or other large AI systems, they all run on servers packed with advanced processors and memory chips. These chips temporarily store huge amounts of data while AI models process requests. As companies race to build larger AI systems, demand for memory has climbed sharply. Technology firms including Google, OpenAI, Anthropic and Meta are investing billions of dollars in AI infrastructure. That has increased competition for memory chips that are also used in consumer products such as laptops, tablets and smartphones.</p><h2><strong>Why Apple is paying more</strong></h2><p>Apple does not manufacture its own memory chips. Instead, it buys DRAM and NAND flash memory from suppliers such as Samsung, SK hynix and Micron. As AI companies place larger and more profitable orders, suppliers have shifted more production towards AI-focused hardware. That has tightened supply for consumer electronics manufacturers. Apple CEO Tim Cook recently described the situation as unlike anything he has seen during his decades in the technology industry. Speaking to The Wall Street Journal, Cook compared the current memory market to a "hundred-year flood" and said suppliers were passing on significant price increases because demand had outpaced available supply. According to Apple, the increase became large enough that the company had to raise prices on selected MacBook and iPad models.</p><h2><strong>Why memory makers are prioritising AI</strong></h2><p>The memory industry has changed dramatically over the past two years. During the slowdown in 2023, manufacturers cut investment after profits fell sharply.</p><p>&nbsp;</p><p>Now, demand has returned but from a different source.</p><p>&nbsp;</p><p>Instead of smartphones and personal computers, AI servers have become the biggest buyers of advanced memory. These customers are often willing to pay more than traditional consumer electronics companies because memory is essential for training and running AI models. That has encouraged suppliers to prioritise AI-related production, further reducing available supply for consumer devices.</p><h2><strong>Could iPhones become more expensive?</strong></h2><p>Apple has not announced price increases for iPhones. However, industry analysts believe the same market pressures could eventually affect other products if memory prices remain elevated.</p><p>Much will depend on whether memory manufacturers expand production quickly enough to meet growing AI demand while maintaining supply for consumer electronics. For now, Apple says it hopes pricing and supply return to more normal levels.</p><h2><strong>AI is changing more than software</strong></h2><p>Most people associate artificial intelligence with chatbots, image generators and digital assistants. But the technology is also reshaping global supply chains. As AI companies continue spending billions on infrastructure, competition for critical components such as memory chips is becoming more intense. That means the AI boom is no longer affecting only technology companies. It is beginning to influence the price people pay for the laptops, tablets and smartphones they use every day.</p>

AI is changing more than just the software people use every day. It is also driving up the cost of the memory chips inside laptops and tablets, forcing Apple to increase prices for some MacBook and iPad models. Here's why it is happening and what it could mean for future devices.

Artificial intelligence is making computers smarter. It is also making them more expensive. Apple recently increased prices for several MacBook and iPad models after memory costs surged, highlighting a growing problem across the technology industry. The company says demand for memory chips has risen so quickly that it has become difficult for consumer electronics makers to secure enough supply at reasonable prices. The shortage is largely being driven by AI companies building massive data centres to train and run increasingly powerful artificial intelligence models. Why AI needs so much memory Every AI model relies on enormous amounts of high-speed memory. Whether it is ChatGPT, Gemini, Claude or other large AI systems, they all run on servers packed with advanced processors and memory chips. These chips temporarily store huge amounts of data while AI models process requests. As companies race to build larger AI systems, demand for memory has climbed sharply. Technology firms including Google, OpenAI, Anthropic and Meta are investing billions of dollars in AI infrastructure. That has increased competition for memory chips that are also used in consumer products such as laptops, tablets and smartphones. Why Apple is paying more Apple does not manufacture its own memory chips. Instead, it buys DRAM and NAND flash memory from suppliers such as Samsung, SK hynix and Micron. As AI companies place larger and more profitable orders, suppliers have shifted more production towards AI-focused hardware. That has tightened supply for consumer electronics manufacturers. Apple CEO Tim Cook recently described the situation as unlike anything he has seen during his decades in the technology industry. Speaking to The Wall Street Journal, Cook compared the current memory market to a "hundred-year flood" and said suppliers were passing on significant price increases because demand had outpaced available supply. According to Apple, the increase became large enough that the company had to raise prices on selected MacBook and iPad models. Why memory makers are prioritising AI The memory industry has changed dramatically over the past two years. During the slowdown in 2023, manufacturers cut investment after profits fell sharply. Now, demand has returned but from a different source. Instead of smartphones and personal computers, AI servers have become the biggest buyers of advanced memory. These customers are often willing to pay more than traditional consumer electronics companies because memory is essential for training and running AI models. That has encouraged suppliers to prioritise AI-related production, further reducing available supply for consumer devices. Could iPhones become more expensive? Apple has not announced price increases for iPhones. However, industry analysts believe the same market pressures could eventually affect other products if memory prices remain elevated. Much will depend on whether memory manufacturers expand production quickly enough to meet growing AI demand while maintaining supply for consumer electronics. For now, Apple says it hopes pricing and supply return to more normal levels. AI is changing more than software Most people associate artificial intelligence with chatbots, image generators and digital assistants. But the technology is also reshaping global supply chains. As AI companies continue spending billions on infrastructure, competition for critical components such as memory chips is becoming more intense. That means the AI boom is no longer affecting only technology companies. It is beginning to influence the price people pay for the laptops, tablets and smartphones they use every day.

<p>AI is changing more than just the software people use every day. It is also driving up the cost of the memory chips inside laptops and tablets, forcing Apple to increase prices for some MacBook and iPad models. Here's why it is happening and what it could mean for future devices.</p>

AI is making your next MacBook more expensive. Here's why

At the 2026 World Economic Forum’s Annual Meeting of the New Champions—widely known as "Summer Davos"—in Dalian, China, the discourse around artificial intelligence shifted from speculative fear to the practical realities of workforce integration. With nearly 40% of jobs currently facing AI-driven transformation, leaders emphasized that the primary challenge is no longer the technology itself, but the "human systems" required to support it.

Debate Continues Over AI's Impact on Human Jobs | Summer Davos

<p>A new open-weight AI model is fast becoming popular, challenging expensive, premium subscriptions like ChatGPT or Claude. A Chinese AI company called Z.ai (formerly Zhipu AI) has released GLM 5.2, which offers frontier-level performance that can be run entirely on local hardware. It goes against a major assumption - that premium subscriptions are the only way to go. ChatGPT and Claude are closed systems, while GLM 5.2 gives developers direct access to the model itself.</p><p>&nbsp;</p><p>Open Source gives you everything, from model weights to training code, a data processing pipeline, an evaluation framework and often the training dataset. Open-source and open-weight AI models are highly flexible and cheaper, but earlier fell behind bigger models. But Meta’s Llama family, Mistral, and GLM 5.2, are now highlighting that not every task needs the most expensive AI model. Not every business needs to solve world-class theoretical logic, and simply needs open models that can complete the majority of their work at a low cost.</p><p>Also Read: <a href="https://www.wionews.com/trending/manual-prompts-are-already-dead-tech-titans-have-moved-on-to-loop-engineering-1782113866510"><strong>Manual prompts are dead. Tech titans have moved on to loop engineering. What is it?</strong></a></p><h2><strong>Challenges with open models</strong></h2><p>GLM 5.2 is a Mixture-of-Experts (MoE) model boasting 744-billion to 753-billion parameters. It features a 1-million-token context window, meaning it can process entire codebases or libraries of books simultaneously. Being an "open-weight" model, it lets developers download its "brain" directly. This allows companies in sensitive industries, like finance or medicine, to keep all data in-house instead of sending it to third-party corporate cloud servers.</p><p>&nbsp;</p><p>However, the open models have severe hardware requirements. Running GLM 5.2 locally comes with intense physical limitations. In its uncompressed form, it takes up a staggering 1.51 terabytes of storage and memory. It needs to be heavily compressed, even if it needs to be loaded on consumer-grade hardware. Even in this case, it demands roughly 240GB of memory just to load. GLM 5.2 might not replace everything, but it shows that open models are becoming cheaper and more competitive.</p>

Chinese AI company called Z.ai (formerly Zhipu AI) has released GLM 5.2, which challenges the idea that you need to pay premium subscription prices to big companies for frontier-level performance. However, open models have their problems.

A new open-weight AI model is fast becoming popular, challenging expensive, premium subscriptions like ChatGPT or Claude. A Chinese AI company called Z.ai (formerly Zhipu AI) has released GLM 5.2, which offers frontier-level performance that can be run entirely on local hardware. It goes against a major assumption - that premium subscriptions are the only way to go. ChatGPT and Claude are closed systems, while GLM 5.2 gives developers direct access to the model itself. Open Source gives you everything, from model weights to training code, a data processing pipeline, an evaluation framework and often the training dataset. Open-source and open-weight AI models are highly flexible and cheaper, but earlier fell behind bigger models. But Meta’s Llama family, Mistral, and GLM 5.2, are now highlighting that not every task needs the most expensive AI model. Not every business needs to solve world-class theoretical logic, and simply needs open models that can complete the majority of their work at a low cost. Also Read: Manual prompts are dead. Tech titans have moved on to loop engineering. What is it? Challenges with open models GLM 5.2 is a Mixture-of-Experts (MoE) model boasting 744-billion to 753-billion parameters. It features a 1-million-token context window, meaning it can process entire codebases or libraries of books simultaneously. Being an "open-weight" model, it lets developers download its "brain" directly. This allows companies in sensitive industries, like finance or medicine, to keep all data in-house instead of sending it to third-party corporate cloud servers. However, the open models have severe hardware requirements. Running GLM 5.2 locally comes with intense physical limitations. In its uncompressed form, it takes up a staggering 1.51 terabytes of storage and memory. It needs to be heavily compressed, even if it needs to be loaded on consumer-grade hardware. Even in this case, it demands roughly 240GB of memory just to load. GLM 5.2 might not replace everything, but it shows that open models are becoming cheaper and more competitive.

<p>Chinese AI company called Z.ai (formerly Zhipu AI) has released GLM 5.2, which challenges the idea that you need to pay premium subscription prices to big companies for frontier-level performance. However, open models have their problems.</p>

Chinese AI which runs locally challenges premium models like ChatGPT and Claude

<p>Meta is making a new push into wearable artificial intelligence. On Tuesday, they unveiled a new line of AI-powered smart glasses developed with eyewear giant EssilorLuxottica. Starting at $299, the new devices are considerably cheaper than Meta's premium smart glasses launched last year.</p><p>&nbsp;</p><p>The announcement is important because it signals Meta's ambition to make AI hardware more accessible to everyday consumers as competition in the wearable technology market intensifies. The launch also comes as rivals including Apple, Google and Snap continue investing heavily in next-generation smart glasses and AI-powered devices.</p><h2><strong>Meta is betting big on AI glasses</strong></h2><p>For several years, Meta has argued that smart glasses could become one of the most important ways people interact with artificial intelligence. Meta’s latest glasses are powered by Meta AI running on Muse Spark, the first model developed by Meta's Superintelligence Labs. According to Meta, the glasses allow users to interact with AI, access information and receive text-based assistance through a wearable device rather than a smartphone. The launch reflects CEO Mark Zuckerberg's broader vision of bringing what he calls "personal intelligence" directly to consumers.</p><figure class="image"><img src="https://about.fb.com/wp-content/uploads/2026/06/02_A-Pair-for-Any-Look_Carousel-02.jpg?fit=1920%2C1672" alt="Meta glasses"></figure><h2><strong>What's different about the new glasses?</strong></h2><p>One notable change is branding. Unlike previous devices, the new glasses are not linked to popular eyewear brands such as Ray-Ban or Oakley. Instead, Meta is introducing its own design language with several new frame styles and colours. Among the options is a slim oval collection developed in collaboration with media personality Kylie Jenner. The company hopes the broader range of styles will help attract consumers who may not have been interested in earlier smart glasses. A growing market</p><figure class="image"><img src="https://about.fb.com/wp-content/uploads/2026/06/02_A-Pair-for-Any-Look_Carousel-03.jpg?fit=1920%2C1672" alt="Meta glasses"></figure><h2><img src="https://about.fb.com/wp-content/uploads/2026/06/02_A-Pair-for-Any-Look_Carousel-01.jpg?fit=1920%2C1672" alt="Meta glasses"><strong>Meta enters the launch from a position of strength</strong></h2><p>According to data from the International Data Corporation (IDC), global smart-glasses shipments reached 9.6 million units last year. Meta accounted for approximately 76.1 per cent of those shipments. That dominance has encouraged competitors to accelerate their own wearable AI projects. Google is expanding its smart-glasses efforts, while Apple is reportedly developing future wearable products that could incorporate artificial intelligence. Meanwhile, Snap recently launched augmented reality glasses priced at $2,195, targeting a different segment of the market.</p><h2><strong>Why the lower price matters</strong></h2><p>The biggest talking point may be the price. At $299, the new Meta Glasses cost significantly less than last year's $800 Ray-Ban Display glasses. Industry analysts have long argued that price remains one of the biggest barriers preventing smart glasses from becoming mainstream. By lowering the entry cost, Meta appears to be prioritising adoption and ecosystem growth over premium positioning.</p><p>&nbsp;</p><p>The next phase of the AI race.Artificial intelligence is increasingly moving beyond chatbots and smartphones. Technology companies are now competing to determine which device will become the primary way people interact with AI throughout the day.</p><h2><strong>Meta believes smart glasses could play that role</strong></h2><p>Whether consumers agree remains to be seen. However, with a lower price, new designs and the company's latest AI technology, Meta is making one thing clear: it sees wearable AI as one of the most important battlegrounds in the technology industry.</p>

Meta has launched a new range of AI-powered smart glasses starting at $299, significantly undercutting previous models. Powered by Meta's latest AI model.

Meta is making a new push into wearable artificial intelligence. On Tuesday, they unveiled a new line of AI-powered smart glasses developed with eyewear giant EssilorLuxottica. Starting at $299, the new devices are considerably cheaper than Meta's premium smart glasses launched last year. The announcement is important because it signals Meta's ambition to make AI hardware more accessible to everyday consumers as competition in the wearable technology market intensifies. The launch also comes as rivals including Apple, Google and Snap continue investing heavily in next-generation smart glasses and AI-powered devices. Meta is betting big on AI glasses For several years, Meta has argued that smart glasses could become one of the most important ways people interact with artificial intelligence. Meta’s latest glasses are powered by Meta AI running on Muse Spark, the first model developed by Meta's Superintelligence Labs. According to Meta, the glasses allow users to interact with AI, access information and receive text-based assistance through a wearable device rather than a smartphone. The launch reflects CEO Mark Zuckerberg's broader vision of bringing what he calls "personal intelligence" directly to consumers. What's different about the new glasses? One notable change is branding. Unlike previous devices, the new glasses are not linked to popular eyewear brands such as Ray-Ban or Oakley. Instead, Meta is introducing its own design language with several new frame styles and colours. Among the options is a slim oval collection developed in collaboration with media personality Kylie Jenner. The company hopes the broader range of styles will help attract consumers who may not have been interested in earlier smart glasses. A growing market Meta enters the launch from a position of strength According to data from the International Data Corporation (IDC), global smart-glasses shipments reached 9.6 million units last year. Meta accounted for approximately 76.1 per cent of those shipments. That dominance has encouraged competitors to accelerate their own wearable AI projects. Google is expanding its smart-glasses efforts, while Apple is reportedly developing future wearable products that could incorporate artificial intelligence. Meanwhile, Snap recently launched augmented reality glasses priced at $2,195, targeting a different segment of the market. Why the lower price matters The biggest talking point may be the price. At $299, the new Meta Glasses cost significantly less than last year's $800 Ray-Ban Display glasses. Industry analysts have long argued that price remains one of the biggest barriers preventing smart glasses from becoming mainstream. By lowering the entry cost, Meta appears to be prioritising adoption and ecosystem growth over premium positioning. The next phase of the AI race.Artificial intelligence is increasingly moving beyond chatbots and smartphones. Technology companies are now competing to determine which device will become the primary way people interact with AI throughout the day. Meta believes smart glasses could play that role Whether consumers agree remains to be seen. However, with a lower price, new designs and the company's latest AI technology, Meta is making one thing clear: it sees wearable AI as one of the most important battlegrounds in the technology industry.

<p>Meta has launched a new range of AI-powered smart glasses starting at $299, significantly undercutting previous models. Powered by Meta's latest AI model.</p>

Meta introduces new AI smart glasses powered by its latest AI model, starting at $299

<p>Loop engineering is quickly replacing the “traditional” practice of giving manual prompts to AI agents, with experts embracing it and pushing for its use. This means that if you are learning to write AI prompts after being told that it is the key to getting the best use out of AI, you are already behind in the race. Boris Cherny, the creator of Claude Code, said in an interview that he no longer writes his own prompts and instead relies on loops. "I don't write the prompt anymore," he said.&nbsp;</p><p>&nbsp;</p><p>So now he directly engages with an enhanced version of Claude, which coordinates the tasks for him. "Claude writes the prompt, and now I'm talking to that new Claude that is kind of coordinating," he added. Loop engineering is also being talked about by OpenAI engineer Peter Steinberger, who created the OpenClaw project.</p><p>Also Read:<a href="https://www.wionews.com/trending/fact-check-did-jeff-bezos-say-human-use-of-water-is-slowing-down-ai-1781967987089"><strong> Fact check: Did Jeff Bezos say human use of water is slowing down AI? Here is what we found</strong></a></p><h2>What is loop engineering?</h2><p>He is telling people to stop writing prompts for AI agents and design loops. So the latest thing in AI is to learn to design these loops that can effectively guide the AI agents. Loops eliminate the need to constantly feed information to AI agents. They are automated systems that repeatedly guide and supervise those agents until the task is over. This is known as loop engineering, where you not only delegate your primary work to AI, but also the act of prompting the instructions.</p><p>Also Read: <a href="https://www.wionews.com/trending/company-ends-up-spending-rs-4-800-crore-on-claude-ai-in-a-month-because-of-a-tiny-oversight-1780048597729"><strong>Company accidentally blows away Rs 4,800 crore on Claude AI in a month because of a tiny oversight</strong></a></p><h2><strong>Manual prompts to AI agents are outdated</strong></h2><p>&nbsp;</p><p>Claire Vo, founder of ChatPRD, and Addy Osmani, Google Cloud's director, have described the method of prompting generative AI tools as "traditional", which is already on its way out. Osmani said, "Loop engineering is replacing yourself as the person who prompts the agent. You design the system that does it instead,” he wrote in an article, adding that manual prompts are "kind of over". He says there are five key elements to creating a loop: automations, worktrees, skills, plugins, and connectors. Vo says many employees might already be employing loops unknowingly, such as the scheduled tasks in ClaudeCowork.</p><p>&nbsp;</p><p>However, AI agents themselves still carry the big question of costs. Companies are assigned a set number of tokens, and operating multiple AI agents means that they burn through them quickly. Recently, an unnamed company burned through Rs. 4,800 crore in a month because it did not set a limit or cap on the use of tokens.&nbsp;</p>

Loop engineering is the latest upgrade in AI and is replacing manual prompts. Experts are telling people to stop giving manual prompts to AI agents and instead move to using loops. This shows the fast advancement in AI and how many of us are already lagging. 

Loop engineering is quickly replacing the “traditional” practice of giving manual prompts to AI agents, with experts embracing it and pushing for its use. This means that if you are learning to write AI prompts after being told that it is the key to getting the best use out of AI, you are already behind in the race. Boris Cherny, the creator of Claude Code, said in an interview that he no longer writes his own prompts and instead relies on loops. "I don't write the prompt anymore," he said. So now he directly engages with an enhanced version of Claude, which coordinates the tasks for him. "Claude writes the prompt, and now I'm talking to that new Claude that is kind of coordinating," he added. Loop engineering is also being talked about by OpenAI engineer Peter Steinberger, who created the OpenClaw project. Also Read: Fact check: Did Jeff Bezos say human use of water is slowing down AI? Here is what we found What is loop engineering? He is telling people to stop writing prompts for AI agents and design loops. So the latest thing in AI is to learn to design these loops that can effectively guide the AI agents. Loops eliminate the need to constantly feed information to AI agents. They are automated systems that repeatedly guide and supervise those agents until the task is over. This is known as loop engineering, where you not only delegate your primary work to AI, but also the act of prompting the instructions. Also Read: Company accidentally blows away Rs 4,800 crore on Claude AI in a month because of a tiny oversight Manual prompts to AI agents are outdated Claire Vo, founder of ChatPRD, and Addy Osmani, Google Cloud's director, have described the method of prompting generative AI tools as "traditional", which is already on its way out. Osmani said, "Loop engineering is replacing yourself as the person who prompts the agent. You design the system that does it instead,” he wrote in an article, adding that manual prompts are "kind of over". He says there are five key elements to creating a loop: automations, worktrees, skills, plugins, and connectors. Vo says many employees might already be employing loops unknowingly, such as the scheduled tasks in ClaudeCowork. However, AI agents themselves still carry the big question of costs. Companies are assigned a set number of tokens, and operating multiple AI agents means that they burn through them quickly. Recently, an unnamed company burned through Rs. 4,800 crore in a month because it did not set a limit or cap on the use of tokens.

<p>Loop engineering is the latest upgrade in AI and is replacing manual prompts. Experts are telling people to stop giving manual prompts to AI agents and instead move to using loops. This shows the fast advancement in AI and how many of us are already lagging.&nbsp;</p>

Manual prompts are dead. Tech titans have moved on to loop engineering. What is it?

<p>Elon Musk believes the future could look very different from today. The world's first trillionaire, whose wealth recently surpassed $1.2 trillion following SpaceX's blockbuster IPO, says artificial intelligence and humanoid robots could eventually create an era of abundance where goods and services become far cheaper and easier to access.</p><p>&nbsp;</p><p>His latest comments came on X, where he responded to a post discussing the future potential of humanoid robots. The original post claimed that millions of advanced robots could rebuild entire cities in months and suggested that billions of such machines could transform the global economy by 2045.</p><p>Musk replied with a short but striking message: “Amazing abundance for all!”</p>

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">Wild times ahead! <a href="https://t.co/kmYjFpaJ3F">https://t.co/kmYjFpaJ3F</a></p>&mdash; Elon Musk (@elonmusk) <a href="https://x.com/elonmusk/status/2068577952875377062?ref_src=twsrc%5Etfw">June 21, 2026</a></blockquote> <script async src="https://platform.x.com/widgets.js" charset="utf-8"></script>

<h2><strong>What does 'amazing abundance' mean?</strong></h2><p>The phrase appears to reflect Musk's long-held belief that AI and robots will dramatically increase productivity. In Musk's vision, machines would handle much of the world's manufacturing, logistics, construction and service work. As a result, products and services could become cheaper because the cost of labour would fall significantly. Rather than focusing on scarcity, Musk believes society could move towards abundance, where many necessities become widely available. His comments have attracted additional attention because Tesla has reportedly filed a trademark application related to the phrase "Amazing Abundance". The move has fuelled speculation that the company may use the term as part of its future messaging around robotics and artificial intelligence.</p><h2><strong>Tesla's Optimus robot is central to the plan</strong></h2><p>At the centre of Musk's vision is Tesla's humanoid robot, Optimus. Tesla has spent several years developing the robot, which is designed to perform repetitive, physically demanding and potentially dangerous tasks. The company believes Optimus could eventually work in factories, warehouses, offices and even homes. Musk has repeatedly argued that humanoid robots could become more significant to Tesla's future than its electric vehicle business. Supporters see Optimus as a potentially transformative product. Critics point out that large-scale deployment remains years away and that significant technical challenges still exist.</p><h2><strong>Why Musk says money may become less important</strong></h2><p>Musk recently expanded on these ideas during the 2026 Abundance Summit. According to Musk, advances in AI and robotics could eventually create so much economic output that traditional ideas about income and work may change. "We're going to have universal high income," Musk said during the event. He argued that AI systems and robots may eventually produce goods and services at such a scale that people will have access to a much higher standard of living regardless of their employment status. Musk even suggested that money could become less relevant in the distant future if abundance becomes widespread. However, he has not provided detailed explanations for how such a system would function economically or politically.</p><h2><strong>The debate over AI and jobs</strong></h2><p>Musk's vision arrives at a time when AI is already reshaping workplaces. Companies across technology, finance and consulting have increasingly adopted AI tools to automate tasks previously handled by employees. Major firms including Amazon, Meta, Oracle and Tata Consultancy Services have announced workforce reductions while simultaneously increasing AI investments. Supporters argue that technological advances have historically created new opportunities even as older jobs disappeared. Others worry that AI could disrupt labour markets faster than societies can adapt.</p><h2><strong>Tesla is not alone in the robot race</strong></h2><p>Tesla may be one of the best-known names in robotics, but it is far from the only company pursuing humanoid machines. Chinese firms such as Unitree continue to showcase increasingly capable robots, while US-based Figure AI is also developing advanced humanoid systems. Competition is intensifying as companies race to build machines capable of performing real-world tasks at scale. Whether Musk's vision of billions of robots becomes reality remains uncertain. But one thing is clear: the world's richest entrepreneur is betting heavily that AI and robotics will play a central role in shaping the economy of the future.</p>

Elon Musk says artificial intelligence and humanoid robots could create "amazing abundance for all" by dramatically increasing productivity and lowering costs. His comments come as Tesla expands its Optimus robot ambitions and AI continues to reshape workplaces around the world.

Elon Musk believes the future could look very different from today. The world's first trillionaire, whose wealth recently surpassed $1.2 trillion following SpaceX's blockbuster IPO, says artificial intelligence and humanoid robots could eventually create an era of abundance where goods and services become far cheaper and easier to access. His latest comments came on X, where he responded to a post discussing the future potential of humanoid robots. The original post claimed that millions of advanced robots could rebuild entire cities in months and suggested that billions of such machines could transform the global economy by 2045. Musk replied with a short but striking message: “Amazing abundance for all!”

<p>Elon Musk says artificial intelligence and humanoid robots could create "amazing abundance for all" by dramatically increasing productivity and lowering costs. His comments come as Tesla expands its Optimus robot ambitions and AI continues to reshape workplaces around the world.</p>

Elon Musk predicts a future where robots make money less important

<p>Viral claims about Jeff Bezos saying that human use of water is delaying the growth of AI are circulating on the internet and social media. However, fact-checking shows that Bezos did not make the controversial statement. The fake claims are being attributed to Bezos, with reports suggesting that he said it at the annual VivaTech conference in Paris a few days ago.&nbsp;</p><h2><strong>Purported statement being attributed to Bezos</strong></h2><p>According to the claim, the billionaire allegedly said, "If we starve our data infrastructure of cooling resources just to sustain baseline human comfort, we're actively delaying the birth of a superintelligence." It further adds, “Sometimes you've to prioritise intelligence that will save us over biology that slows us.”</p><h2><strong>Fact check: No, he did not say anything about water and human use</strong></h2><p>&nbsp;</p><p>He is being slammed for the purported words, but the truth is that Bezos did not say any of this. The unedited clip from the event proves that the words are wrongly being attributed to Bezos. We went through the live stream of the VivaTech event posted by Associated Press and did not come across the contentious words. Other people on the internet also wrote similar posts, saying that Bezos did not make any such statement. The billionaire talked about his space company, Blue Origin and his industrial AI startup, Prometheus. However, he did not talk about the use of water for data centres.</p><h2><strong>Reddit thread deletes post</strong></h2><p>&nbsp;</p><p>Another check shows that the fake news originated from a Reddit thread, which has now deleted the post. Examining the statement on Gemini also revealed that it is incorrect. "This quote is entirely fake, though it recently circulated on platforms like Reddit, Hacker News, and X (formerly Twitter) wrapped in a convincing, manufactured screenshot," the AI tool responded when asked about the authenticity of the statement.</p><p>&nbsp;</p><figure class="image"><img src="https://cdn1.wionews.com/prod/wion/images/2026/20260620/image-1781968077842.png"></figure><h2><strong>What Bezos said about AI at VivaTech conference</strong></h2><p>&nbsp;</p><p>At the VivaTech conference, Bezos talked about Blue Origin and its role in establishing future moon bases. He also talked about Prometheus and AI in general. Bezos said that AI will not reduce jobs, but create a shortage of labour. "AI is going to create a labour shortage because it’s going to make it possible for people to identify more problems,” Bezos said during a session moderated by former NASA astronaut Mike Massimino. Dave Limp, chief executive of Blue Origin, was also at the event. Bezos goes on to say that the only thing limiting growth will be imagination.<br>&nbsp;</p>

Amazon founder Jeff Bezos is going viral for allegedly saying that human use of water is slowing down the growth of AI. However, it is a fake statement and completely fabricated. We heard the conference and found that Bezos did not say it.

Viral claims about Jeff Bezos saying that human use of water is delaying the growth of AI are circulating on the internet and social media. However, fact-checking shows that Bezos did not make the controversial statement. The fake claims are being attributed to Bezos, with reports suggesting that he said it at the annual VivaTech conference in Paris a few days ago. Purported statement being attributed to Bezos According to the claim, the billionaire allegedly said, "If we starve our data infrastructure of cooling resources just to sustain baseline human comfort, we're actively delaying the birth of a superintelligence." It further adds, “Sometimes you've to prioritise intelligence that will save us over biology that slows us.” Fact check: No, he did not say anything about water and human use He is being slammed for the purported words, but the truth is that Bezos did not say any of this. The unedited clip from the event proves that the words are wrongly being attributed to Bezos. We went through the live stream of the VivaTech event posted by Associated Press and did not come across the contentious words. Other people on the internet also wrote similar posts, saying that Bezos did not make any such statement. The billionaire talked about his space company, Blue Origin and his industrial AI startup, Prometheus. However, he did not talk about the use of water for data centres. Reddit thread deletes post Another check shows that the fake news originated from a Reddit thread, which has now deleted the post. Examining the statement on Gemini also revealed that it is incorrect. "This quote is entirely fake, though it recently circulated on platforms like Reddit, Hacker News, and X (formerly Twitter) wrapped in a convincing, manufactured screenshot," the AI tool responded when asked about the authenticity of the statement. What Bezos said about AI at VivaTech conference At the VivaTech conference, Bezos talked about Blue Origin and its role in establishing future moon bases. He also talked about Prometheus and AI in general. Bezos said that AI will not reduce jobs, but create a shortage of labour. "AI is going to create a labour shortage because it’s going to make it possible for people to identify more problems,” Bezos said during a session moderated by former NASA astronaut Mike Massimino. Dave Limp, chief executive of Blue Origin, was also at the event. Bezos goes on to say that the only thing limiting growth will be imagination.

<p>Amazon founder Jeff Bezos is going viral for allegedly saying that human use of water is slowing down the growth of AI. However, it is a fake statement and completely fabricated. We heard the conference and found that Bezos did not say it.</p>

Fact check: Did Jeff Bezos say human use of water is slowing down AI? Here is what we found

<p>As companies around the world rush to integrate artificial intelligence into daily operations, one CEO has shared an example of how difficult that transition can become. Eric Vaughan, chief executive of software company IgniteTech, says the firm replaced nearly 80 per cent of its workforce after many employees refused to adopt AI tools despite extensive training and company-funded support.</p><p>&nbsp;</p><p>The comments, first reported by Fortune, have attracted attention because they illustrate a growing challenge facing businesses: technology can be introduced quickly, but changing workplace habits may take much longer. According to Vaughan, resistance to AI became one of the biggest obstacles in the company's transformation into what he describes as an AI-first organisation.</p><h2><strong>The company spent heavily on AI training</strong></h2><p>Vaughan said IgniteTech invested significant resources into helping employees learn how to use artificial intelligence. The company introduced weekly "AI Mondays", where workers focused exclusively on AI-related projects. Employees were also offered reimbursements for AI tools, prompt-engineering courses and external training programmes. The company even brought in outside experts to support staff. According to Vaughan, IgniteTech dedicated around 20 per cent of its payroll towards AI learning efforts for one day each week over an entire quarter. Despite that investment, some employees remained unwilling to embrace the technology. "We took 20 per cent of our entire payroll for one day a week for an entire quarter to get people to learn AI and still found people that said, 'I'm not going to do it,'" Vaughan said.</p><h2><strong>Why the company replaced so many employees</strong></h2><p>The transition began in early 2023 after Vaughan became convinced that generative AI represented a major shift for businesses. He viewed AI as a technology that companies would eventually need to adopt to remain competitive. As a result, IgniteTech pushed aggressively towards integrating AI into products, workflows and employee responsibilities. However, Vaughan said some workers rejected the direction entirely. Between 2023 and early 2024, the company replaced nearly 80 per cent of its workforce, affecting hundreds of employees. "That was not our goal," Vaughan said. "But changing minds was harder than adding skills."</p><h2><strong>A wider debate over AI in the workplace</strong></h2><p>The comments arrive at a time when artificial intelligence is transforming industries ranging from software development and customer support to finance and healthcare. Many employers are encouraging workers to use AI tools to improve productivity and automate repetitive tasks. At the same time, concerns remain about job security, changing workplace expectations and the speed of technological change. IgniteTech's experience highlights a growing question facing employers: how should organisations respond when employees are unwilling to adopt technologies that leadership views as essential?</p><h2><strong>Vaughan says he would make the same decision again</strong></h2><p>While defending the outcome, Vaughan described the process as difficult. He said he would make the same decision again but stopped short of recommending that other businesses follow the same approach. According to the CEO, alignment between leadership and employees ultimately mattered more than training programmes alone. "If your team isn't in sync with you, I don't think any amount of training or strategy is going to fix that problem," he said.</p><h2><strong>What this means for workers and businesses</strong></h2><p>The story reflects a broader shift taking place across the technology industry. As AI tools become more capable, employers are increasingly expecting workers to understand and use them as part of their roles. For employees, AI skills are rapidly becoming a valuable workplace asset. For businesses, the challenge is not only introducing new technology but also helping teams adapt to it. The debate is likely to continue as artificial intelligence becomes more deeply embedded in everyday work.</p><p>&nbsp;</p><p>&nbsp;</p>

A Texas-based software company replaced nearly 80 per cent of its workforce after many employees resisted adopting artificial intelligence, according to CEO Eric Vaughan. 

As companies around the world rush to integrate artificial intelligence into daily operations, one CEO has shared an example of how difficult that transition can become. Eric Vaughan, chief executive of software company IgniteTech, says the firm replaced nearly 80 per cent of its workforce after many employees refused to adopt AI tools despite extensive training and company-funded support. The comments, first reported by Fortune, have attracted attention because they illustrate a growing challenge facing businesses: technology can be introduced quickly, but changing workplace habits may take much longer. According to Vaughan, resistance to AI became one of the biggest obstacles in the company's transformation into what he describes as an AI-first organisation. The company spent heavily on AI training Vaughan said IgniteTech invested significant resources into helping employees learn how to use artificial intelligence. The company introduced weekly "AI Mondays", where workers focused exclusively on AI-related projects. Employees were also offered reimbursements for AI tools, prompt-engineering courses and external training programmes. The company even brought in outside experts to support staff. According to Vaughan, IgniteTech dedicated around 20 per cent of its payroll towards AI learning efforts for one day each week over an entire quarter. Despite that investment, some employees remained unwilling to embrace the technology. "We took 20 per cent of our entire payroll for one day a week for an entire quarter to get people to learn AI and still found people that said, 'I'm not going to do it,'" Vaughan said. Why the company replaced so many employees The transition began in early 2023 after Vaughan became convinced that generative AI represented a major shift for businesses. He viewed AI as a technology that companies would eventually need to adopt to remain competitive. As a result, IgniteTech pushed aggressively towards integrating AI into products, workflows and employee responsibilities. However, Vaughan said some workers rejected the direction entirely. Between 2023 and early 2024, the company replaced nearly 80 per cent of its workforce, affecting hundreds of employees. "That was not our goal," Vaughan said. "But changing minds was harder than adding skills." A wider debate over AI in the workplace The comments arrive at a time when artificial intelligence is transforming industries ranging from software development and customer support to finance and healthcare. Many employers are encouraging workers to use AI tools to improve productivity and automate repetitive tasks. At the same time, concerns remain about job security, changing workplace expectations and the speed of technological change. IgniteTech's experience highlights a growing question facing employers: how should organisations respond when employees are unwilling to adopt technologies that leadership views as essential? Vaughan says he would make the same decision again While defending the outcome, Vaughan described the process as difficult. He said he would make the same decision again but stopped short of recommending that other businesses follow the same approach. According to the CEO, alignment between leadership and employees ultimately mattered more than training programmes alone. "If your team isn't in sync with you, I don't think any amount of training or strategy is going to fix that problem," he said. What this means for workers and businesses The story reflects a broader shift taking place across the technology industry. As AI tools become more capable, employers are increasingly expecting workers to understand and use them as part of their roles. For employees, AI skills are rapidly becoming a valuable workplace asset. For businesses, the challenge is not only introducing new technology but also helping teams adapt to it. The debate is likely to continue as artificial intelligence becomes more deeply embedded in everyday work.

<p>A Texas-based software company replaced nearly 80 per cent of its workforce after many employees resisted adopting artificial intelligence, according to CEO Eric Vaughan.&nbsp;</p>

‘Fired them all, will do again’: Tech CEO replaced 80% of employees with AI, says humans not ready to embrace technology

The U.S. Department of Justice has declared a private AI company’s data center a national security asset and then used that designation to attempt to dismiss a civil rights lawsuit, signaling a growing clash between Silicon Valley power and regulatory accountability.

US Memphis Air Rights vs Pentagon AI Needs

<p>A major shift is coming globally as Pluto is now firmly embedded in Aquarius and will stay there till 2044. This transition, which started on November 19, 2024, is not ordinary and will reshape how society organises power, knowledge, and human connection. This is one of the most significant astrological shifts of the century. Pluto is moving from the earth-bound Capricorn to the air-driven, technological realm of Aquarius. It moves slowly and so remains in one sign for 11 to 30 years. This leads it to shape entire generations, astrologer Harshada Desai says.</p><p>&nbsp;</p><p>In Roman mythology, Pluto was the god of the underworld, the ruler of everything hidden beneath the surface. It carries that same energy in astrology as well. It is a planet of total transformation and is now moving through the sign of collective power and technology, which will spell massive changes for the world. "It is the planet of death and rebirth, of tearing down what is broken so something better can grow in its place," she said. However, this transformation will not be gentle.</p><p>&nbsp;</p><p>"Pluto exposes, dismantles, and transforms — whether we are ready or not. Wherever Pluto travels, it asks one uncompromising question: what needs to die so something truer can live?"</p><p>Pluto in Aquarius will affect entire groups on a larger scale - communities, networks, technology, innovation, and humanitarianism. Most other zodiac signs focus on the individual, Aquarius thinks in systems and masses: how do we organise society? How do we share knowledge? How do we make things work for everyone, not just the few? Harshada says.</p><p>Also Read: <a href="https://www.wionews.com/trending/jupiter-s-exaltation-in-cancer-time-for-emotional-healing-and-finding-soul-s-true-purpose-1781248198931"><strong>Jupiter’s exaltation in Cancer: Time for emotional healing and finding soul's true purpose</strong></a></p><h2><strong>Shift in power</strong></h2><p>The implications of this movement will include the fundamental reshaping of who holds power in society and in what manner. Networks, communities, and decentralised systems will challenge old hierarchies and institutions. The entire structure of the society could change, but it will happen in phases throughout this long period.</p><p>&nbsp;</p><p>The technology element will see a huge surge, which today we can directly link to Artificial Intelligence. It is here and growing, and this phase of Pluto in Aquarius will lead to an upheaval. It will be embedded in infrastructure - healthcare, legal systems, education, and governance. Meanwhile, the debate around who benefits from AI will continue to grow.</p><p>Also Read: <a href="https://www.wionews.com/trending/astrological-significance-of-blue-moon-2026-which-zodiac-signs-and-date-of-births-will-it-affect-1780131894994"><strong>Astrological significance of Blue Moon 2026: Which zodiac signs and date of births will it affect</strong></a></p><h2><strong>Communities over individuals</strong></h2><p>Power will keep moving away from central authorities towards distributed communities. Be it finance, governance or media. From individual we will move to community, as I becomes We.</p><p>Local and digital systems will influence policies, as neighbourhood networks, mutual aid groups, and issue-specific collectives will achieve outcomes that traditional political channels have stalled on, Harshada says.</p><p>&nbsp;</p><p>The scientific field will continue to advance with breakthroughs expected in climate technology, public health, and renewable energy, and will increasingly be framed as collective rights, not consumer products.</p><p>&nbsp;</p><p>Harshada says the primary takeaway of this transition is that power will not remain concentrated at the top.</p>

Pluto has firmly planted itself in Aquarius and will remain there for the next 18 years. This period will see the rise of communities and technology and bring about major shifts on the community level. Pluto, combined with Aquarius, dictates major shifts.

A major shift is coming globally as Pluto is now firmly embedded in Aquarius and will stay there till 2044. This transition, which started on November 19, 2024, is not ordinary and will reshape how society organises power, knowledge, and human connection. This is one of the most significant astrological shifts of the century. Pluto is moving from the earth-bound Capricorn to the air-driven, technological realm of Aquarius. It moves slowly and so remains in one sign for 11 to 30 years. This leads it to shape entire generations, astrologer Harshada Desai says. In Roman mythology, Pluto was the god of the underworld, the ruler of everything hidden beneath the surface. It carries that same energy in astrology as well. It is a planet of total transformation and is now moving through the sign of collective power and technology, which will spell massive changes for the world. "It is the planet of death and rebirth, of tearing down what is broken so something better can grow in its place," she said. However, this transformation will not be gentle. "Pluto exposes, dismantles, and transforms — whether we are ready or not. Wherever Pluto travels, it asks one uncompromising question: what needs to die so something truer can live?" Pluto in Aquarius will affect entire groups on a larger scale - communities, networks, technology, innovation, and humanitarianism. Most other zodiac signs focus on the individual, Aquarius thinks in systems and masses: how do we organise society? How do we share knowledge? How do we make things work for everyone, not just the few? Harshada says. Also Read: Jupiter’s exaltation in Cancer: Time for emotional healing and finding soul's true purpose Shift in power The implications of this movement will include the fundamental reshaping of who holds power in society and in what manner. Networks, communities, and decentralised systems will challenge old hierarchies and institutions. The entire structure of the society could change, but it will happen in phases throughout this long period. The technology element will see a huge surge, which today we can directly link to Artificial Intelligence. It is here and growing, and this phase of Pluto in Aquarius will lead to an upheaval. It will be embedded in infrastructure - healthcare, legal systems, education, and governance. Meanwhile, the debate around who benefits from AI will continue to grow. Also Read: Astrological significance of Blue Moon 2026: Which zodiac signs and date of births will it affect Communities over individuals Power will keep moving away from central authorities towards distributed communities. Be it finance, governance or media. From individual we will move to community, as I becomes We. Local and digital systems will influence policies, as neighbourhood networks, mutual aid groups, and issue-specific collectives will achieve outcomes that traditional political channels have stalled on, Harshada says. The scientific field will continue to advance with breakthroughs expected in climate technology, public health, and renewable energy, and will increasingly be framed as collective rights, not consumer products. Harshada says the primary takeaway of this transition is that power will not remain concentrated at the top.

<p>Pluto has firmly planted itself in Aquarius and will remain there for the next 18 years. This period will see the rise of communities and technology and bring about major shifts on the community level. Pluto, combined with Aquarius, dictates major shifts.</p>

Astrology has an explanation for why AI has become a mammoth

<p>Artificial Intelligence is at the centre of conversations at the G7 Summit in Évian-les-Bains, France, with member countries discussing access to cutting-edge models. There is a scramble for new international "deals" and frameworks as G7 leaders reportedly talked about a “trusted partners” plan where selective access to advanced AI models from US firms such as Anthropic is granted. The matter assumed high priority after Anthropic disabled access for all users to Fable 5 and Mythos 5 on Friday. This was done after US President Donald Trump ordered Anthropic to block foreign nationals from accessing its most advanced AI models. He said that it had to be done due to cybersecurity and national security concerns.</p><p>&nbsp;</p><p>Financial Times reported that a US delegation led by Commerce Secretary Howard Lutnick and European diplomats is holding discussions to have a system where close, trusted G7 allies, either companies or countries, would be able to access American AI models that are currently not open to export. Such an agreement would allow G7 countries to use the AI technology to strengthen cybersecurity defences against rivals such as China.</p><h2><strong>Anthropic's Fable 5 and Mythos 5 models</strong></h2><p>&nbsp;</p><p>Fable 5 and Mythos 5 are yet to be released models by Anthropic. Cybersecurity experts have raised concerns about Mythos, which has been designed to find flaws in computer code. They believe that it may lead to a dramatic escalation in attacks on banks' technology systems. An export control directive on Friday told Anthropic to suspend access to its latest AI models, Fable 5 and Mythos 5, "by any foreign national, whether inside or outside the United States." Both models were released just three days before the order was issued.</p><p>Also Read: <a href="https://www.wionews.com/world/ai-export-controls-intelligence-as-commodity-digital-colonialism-1781617339741"><strong>AI, data centres and digital colonialism: How intelligence became a commodity</strong></a></p><p>&nbsp;</p><p>Countries are now seeking access to Mythos to study how the model could impact safety and find kinks in the computer systems. Before Trump's order, select organisations in "more than 15 countries" were granted access to Mythos so they could use it to detect vulnerabilities in their computer systems.</p><p>Also Read: <a href="https://www.wionews.com/trending/company-ends-up-spending-rs-4-800-crore-on-claude-ai-in-a-month-because-of-a-tiny-oversight-1780048597729"><strong>Company accidentally blows away Rs 4,800 crore on Claude AI in a month because of a tiny oversight</strong></a></p><h2><strong>Technology independence debate at G7</strong></h2><p><br>The export restrictions have led to debate about freedom of technology and caused noticeable friction among G7 members. Henna Virkkunen, the European Commission’s tech chief, has called for removing the "discriminatory" tech measures. However, it has come as a wake-up call for European leaders who are pushing for quick funding of the domestic European AI alternatives. Earlier, Canadian Prime Minister Mark Carney said that the export restrictions highlight the dangers of overreliance on a limited number of American providers.</p><p>Also Read: <a href="https://www.wionews.com/india-news/modi-meets-trump-on-g7-sidelines-in-first-in-person-interaction-since-2025-1781620931627"><strong>WATCH | Modi meets Trump on G7 sidelines in first in-person interaction since 2025</strong></a></p><h2><strong>AI leaders meet for formal working lunch</strong></h2><p><br>Notably, AI executives from Anthropic, OpenAI, and Google, are in France and scheduled to attend a working lunch on Wednesday to address issues like regulation, AI infrastructure, and networks. In an unprecedented move for a G7 summit, the world’s top AI executives are gathering for sessions. The main stage working lunch will include OpenAI CEO Sam Altman, Google DeepMind's Demis Hassabis, Anthropic’s Dario Amodei, and Mistral AI’s Arthur Mensch, alongside executives from Meta, Salesforce, and Cohere. They will be discussing pointers on AI infrastructure, especially its energy demands and the sustainability of the data centres.</p><p>&nbsp;</p><h2><strong>AI safety at G7</strong></h2><p><br>Two other points on the table include protecting minors and trust. French President Emmanuel Macron has led an agenda item of a joint declaration aimed at mitigating the risks of AI-generated harmful content for minors. &nbsp;Other member nations, including Indian Prime Minister Narendra Modi, are calling for "humanity-first" transparency and trust frameworks, so AI deployment closes economic productivity gaps and does not widen them.</p>

Artificial Intelligence at G7: World leaders at the summit in France are actively discussing AI, particularly the export ban the US placed on Anthropic regarding its new AI models, Fable 5 and Mythos 5. Access to technology is important to detect flaws in systems.

Artificial Intelligence is at the centre of conversations at the G7 Summit in Évian-les-Bains, France, with member countries discussing access to cutting-edge models. There is a scramble for new international "deals" and frameworks as G7 leaders reportedly talked about a “trusted partners” plan where selective access to advanced AI models from US firms such as Anthropic is granted. The matter assumed high priority after Anthropic disabled access for all users to Fable 5 and Mythos 5 on Friday. This was done after US President Donald Trump ordered Anthropic to block foreign nationals from accessing its most advanced AI models. He said that it had to be done due to cybersecurity and national security concerns. Financial Times reported that a US delegation led by Commerce Secretary Howard Lutnick and European diplomats is holding discussions to have a system where close, trusted G7 allies, either companies or countries, would be able to access American AI models that are currently not open to export. Such an agreement would allow G7 countries to use the AI technology to strengthen cybersecurity defences against rivals such as China. Anthropic's Fable 5 and Mythos 5 models Fable 5 and Mythos 5 are yet to be released models by Anthropic. Cybersecurity experts have raised concerns about Mythos, which has been designed to find flaws in computer code. They believe that it may lead to a dramatic escalation in attacks on banks' technology systems. An export control directive on Friday told Anthropic to suspend access to its latest AI models, Fable 5 and Mythos 5, "by any foreign national, whether inside or outside the United States." Both models were released just three days before the order was issued. Also Read: AI, data centres and digital colonialism: How intelligence became a commodity Countries are now seeking access to Mythos to study how the model could impact safety and find kinks in the computer systems. Before Trump's order, select organisations in "more than 15 countries" were granted access to Mythos so they could use it to detect vulnerabilities in their computer systems. Also Read: Company accidentally blows away Rs 4,800 crore on Claude AI in a month because of a tiny oversight Technology independence debate at G7 The export restrictions have led to debate about freedom of technology and caused noticeable friction among G7 members. Henna Virkkunen, the European Commission’s tech chief, has called for removing the "discriminatory" tech measures. However, it has come as a wake-up call for European leaders who are pushing for quick funding of the domestic European AI alternatives. Earlier, Canadian Prime Minister Mark Carney said that the export restrictions highlight the dangers of overreliance on a limited number of American providers. Also Read: WATCH | Modi meets Trump on G7 sidelines in first in-person interaction since 2025 AI leaders meet for formal working lunch Notably, AI executives from Anthropic, OpenAI, and Google, are in France and scheduled to attend a working lunch on Wednesday to address issues like regulation, AI infrastructure, and networks. In an unprecedented move for a G7 summit, the world’s top AI executives are gathering for sessions. The main stage working lunch will include OpenAI CEO Sam Altman, Google DeepMind's Demis Hassabis, Anthropic’s Dario Amodei, and Mistral AI’s Arthur Mensch, alongside executives from Meta, Salesforce, and Cohere. They will be discussing pointers on AI infrastructure, especially its energy demands and the sustainability of the data centres. AI safety at G7 Two other points on the table include protecting minors and trust. French President Emmanuel Macron has led an agenda item of a joint declaration aimed at mitigating the risks of AI-generated harmful content for minors. Other member nations, including Indian Prime Minister Narendra Modi, are calling for "humanity-first" transparency and trust frameworks, so AI deployment closes economic productivity gaps and does not widen them.

<p>Artificial Intelligence at G7: World leaders at the summit in France are actively discussing AI, particularly the export ban the US placed on Anthropic regarding its new AI models, Fable 5 and Mythos 5. Access to technology is important to detect flaws in systems.</p>

AI at G7: 'Trusted partners' plan being discussed amid US export ban on Anthropic's AI models

<p>A major shift is quietly taking place online. For years, finding information on the internet followed a familiar pattern. People opened a search engine, typed a few keywords and clicked through a list of links.</p><p>&nbsp;</p><p><strong>That model is now changing</strong></p><p>Increasingly, users are turning to AI assistants such as ChatGPT, Gemini, Claude and Perplexity to get answers directly rather than browsing multiple websites. The change may seem subtle today, but it could become one of the most significant transformations in the history of the internet.</p><p>From searching for links to receiving answers</p><p>Traditional search engines were built around keywords. A user searching for "best laptop for students" or "how to invest money" would receive pages of links and decide which websites to visit.</p><h2><strong>AI systems work differently.</strong></h2><p>Instead of presenting a list of results, they often generate a direct response by analysing information from multiple sources. For users, this can save time.</p><p>Instead of opening five or ten websites, many people now receive a summary in seconds. This shift is becoming increasingly visible through products such as Google's AI Overviews and AI-powered search tools.</p><h2><strong>Why this matters for websites and publishers</strong></h2><p>The consequences extend beyond how people search. For years, websites focused heavily on ranking highly in search results. Businesses invested in search engine optimisation, while publishers competed for clicks from search traffic. Now, visibility increasingly depends on whether AI systems choose to reference, summarise or recommend content. In simple terms, the challenge is no longer just appearing in search results. It is becoming about being recognised as a trusted source that AI systems can understand and cite. This is encouraging publishers and businesses to focus more on expertise, original reporting, strong branding and clear information.</p><h2><strong>AI is becoming a new discovery layer</strong></h2><p>Many technology experts describe the current transition as a move from keyword-driven search to AI-mediated discovery. In the past, people actively searched for information.&nbsp; Now information is increasingly being surfaced, recommended or summarised by algorithms and AI assistants. This trend can already be seen across social media, recommendation systems and AI chat platforms.</p><p>The result is a new discovery layer sitting between users and the wider internet. Instead of directly navigating the web, many people are beginning to rely on AI systems to guide them.</p><p>&nbsp;</p><p>The shift does not mean traditional search engines will disappear. Search remains one of the most important services on the internet and continues to process billions of queries every day. However, the balance is changing. Future online experiences may involve conversations rather than keyword searches. Users may increasingly ask questions, request recommendations or seek explanations from AI systems that understand context and intent. For businesses, creators and publishers, this could require a new approach to visibility. Success may depend less on ranking for individual keywords and more on producing useful, trustworthy and authoritative information that both humans and AI systems can understand.</p><h2><strong>The internet's next chapter</strong></h2><p>Every era of the internet has been defined by a dominant way of finding information. First came directories. Then search engines. Later, social media feeds became powerful discovery tools. Now AI appears to be creating another shift. Most people still think of the internet as a place where they search for information. Increasingly, however, information is being selected, summarised and presented by AI systems before users ever reach a website. The transition is still unfolding, but its impact is already being felt across technology, media, business and everyday online life.</p>

A major change is reshaping the internet as users move from keyword-based searches to AI-powered discovery. Tools such as ChatGPT, Gemini and Perplexity are increasingly delivering answers directly, altering how people find information and how websites gain visibility.


A major shift is quietly taking place online. For years, finding information on the internet followed a familiar pattern. People opened a search engine, typed a few keywords and clicked through a list of links. That model is now changing Increasingly, users are turning to AI assistants such as ChatGPT, Gemini, Claude and Perplexity to get answers directly rather than browsing multiple websites. The change may seem subtle today, but it could become one of the most significant transformations in the history of the internet. From searching for links to receiving answers Traditional search engines were built around keywords. A user searching for "best laptop for students" or "how to invest money" would receive pages of links and decide which websites to visit. AI systems work differently. Instead of presenting a list of results, they often generate a direct response by analysing information from multiple sources. For users, this can save time. Instead of opening five or ten websites, many people now receive a summary in seconds. This shift is becoming increasingly visible through products such as Google's AI Overviews and AI-powered search tools. Why this matters for websites and publishers The consequences extend beyond how people search. For years, websites focused heavily on ranking highly in search results. Businesses invested in search engine optimisation, while publishers competed for clicks from search traffic. Now, visibility increasingly depends on whether AI systems choose to reference, summarise or recommend content. In simple terms, the challenge is no longer just appearing in search results. It is becoming about being recognised as a trusted source that AI systems can understand and cite. This is encouraging publishers and businesses to focus more on expertise, original reporting, strong branding and clear information. AI is becoming a new discovery layer Many technology experts describe the current transition as a move from keyword-driven search to AI-mediated discovery. In the past, people actively searched for information. Now information is increasingly being surfaced, recommended or summarised by algorithms and AI assistants. This trend can already be seen across social media, recommendation systems and AI chat platforms. The result is a new discovery layer sitting between users and the wider internet. Instead of directly navigating the web, many people are beginning to rely on AI systems to guide them. The shift does not mean traditional search engines will disappear. Search remains one of the most important services on the internet and continues to process billions of queries every day. However, the balance is changing. Future online experiences may involve conversations rather than keyword searches. Users may increasingly ask questions, request recommendations or seek explanations from AI systems that understand context and intent. For businesses, creators and publishers, this could require a new approach to visibility. Success may depend less on ranking for individual keywords and more on producing useful, trustworthy and authoritative information that both humans and AI systems can understand. The internet's next chapter Every era of the internet has been defined by a dominant way of finding information. First came directories. Then search engines. Later, social media feeds became powerful discovery tools. Now AI appears to be creating another shift. Most people still think of the internet as a place where they search for information. Increasingly, however, information is being selected, summarised and presented by AI systems before users ever reach a website. The transition is still unfolding, but its impact is already being felt across technology, media, business and everyday online life.

<p>A major change is reshaping the internet as users move from keyword-based searches to AI-powered discovery. Tools such as ChatGPT, Gemini and Perplexity are increasingly delivering answers directly, altering how people find information and how websites gain visibility.</p>

A massive shift is happening online, and most people haven't noticed

<p>Artificial Intelligence systems might have solved an 80-year-old Math problem, but a test designed to evaluate it against top-tier academicians has revealed that it still has a long way to go when it comes to new and complex research-level mathematics. Four AI systems failed the test, which is part of "First Proof," a rigorous new AI benchmarking mechanism. The AI systems were given 10 research-level maths problems, and their answers were assessed by a jury of anonymous human specialists. To prevent AI from simply scanning through its training data, all the questions were new and had never been published anywhere. They were provided by 10 researchers from a broad range of mathematical specialities who had solved them at some point in their careers. A group of 30 mathematicians checked the answers, and the results proved that AI does not come close to human mathematicians in novel problems. The findings were published in the Nature journal.</p><p>Also Read: <a href="https://www.wionews.com/trending/company-ends-up-spending-rs-4-800-crore-on-claude-ai-in-a-month-because-of-a-tiny-oversight-1780048597729"><strong>Company accidentally blows away Rs 4,800 crore on Claude AI in a month because of a tiny oversight</strong></a></p><p>&nbsp;</p><p>The First Proof team carried out the test in an entirely autonomous way without humans being involved in any way. The AI systems also had to be publicly available. This ruled out Google’s Aletheia, which has been designed specifically for solving maths problems, and the full, unreleased version of Claude Mythos. OpenAI entered its ChatGPT 5.5 Pro model. Three academic groups provided the rest of the models - from the University of California, Los Angeles (UCLA); Princeton University in New Jersey; and the Swiss Federal Institute of Technology (ETH) in Zurich. They built "harnesses"—automated systems that prompted existing chatbots, such as ChatGPT, Gemini, and Claude, to repeatedly check the answers.</p><p>Also Read: <a href="https://www.wionews.com/technology/openai-s-ai-model-solves-decades-old-maths-problem-proof-verified-by-researchers-1780210625561"><strong>OpenAI’s AI model solves decades-old maths problem, proof verified by researchers</strong></a></p><h2><strong>AI vs human mathematicians: The results</strong></h2><p><br>The top-performing AI system was able to solve only 6 out of 10 problems. Since all of them had been solved by mathematicians, this proved that human expertise in the field still outperforms AI on unseen, highly rigorous research problems. ETH Zurich stood first, whose harness had used ChatGPT to generate answers. UCLA was second (ChatGPT harness), followed by OpenAI (standalone ChatGPT 5.5 Pro), and Princeton (using a Gemini 3.1 Pro harness).</p><p>The results show that AI models have a long way to go before they can function autonomously as reliable research assistants, proof-checkers, and problem-solvers for human mathematicians. First Proof will come up with more versions of the test in future to correctly decide when this could actually become possible.</p>

AI failed to beat humans in 10 Math problems that expert mathematicians had solved in the past. Four systems entered a test, which was assessed by 30 analysts, and none of them was able to solve all 10 problems. The finding shows humans are still better than AI on new Math problems.

Artificial Intelligence systems might have solved an 80-year-old Math problem, but a test designed to evaluate it against top-tier academicians has revealed that it still has a long way to go when it comes to new and complex research-level mathematics. Four AI systems failed the test, which is part of "First Proof," a rigorous new AI benchmarking mechanism. The AI systems were given 10 research-level maths problems, and their answers were assessed by a jury of anonymous human specialists. To prevent AI from simply scanning through its training data, all the questions were new and had never been published anywhere. They were provided by 10 researchers from a broad range of mathematical specialities who had solved them at some point in their careers. A group of 30 mathematicians checked the answers, and the results proved that AI does not come close to human mathematicians in novel problems. The findings were published in the Nature journal. Also Read: Company accidentally blows away Rs 4,800 crore on Claude AI in a month because of a tiny oversight The First Proof team carried out the test in an entirely autonomous way without humans being involved in any way. The AI systems also had to be publicly available. This ruled out Google’s Aletheia, which has been designed specifically for solving maths problems, and the full, unreleased version of Claude Mythos. OpenAI entered its ChatGPT 5.5 Pro model. Three academic groups provided the rest of the models - from the University of California, Los Angeles (UCLA); Princeton University in New Jersey; and the Swiss Federal Institute of Technology (ETH) in Zurich. They built "harnesses"—automated systems that prompted existing chatbots, such as ChatGPT, Gemini, and Claude, to repeatedly check the answers. Also Read: OpenAI’s AI model solves decades-old maths problem, proof verified by researchers AI vs human mathematicians: The results The top-performing AI system was able to solve only 6 out of 10 problems. Since all of them had been solved by mathematicians, this proved that human expertise in the field still outperforms AI on unseen, highly rigorous research problems. ETH Zurich stood first, whose harness had used ChatGPT to generate answers. UCLA was second (ChatGPT harness), followed by OpenAI (standalone ChatGPT 5.5 Pro), and Princeton (using a Gemini 3.1 Pro harness). The results show that AI models have a long way to go before they can function autonomously as reliable research assistants, proof-checkers, and problem-solvers for human mathematicians. First Proof will come up with more versions of the test in future to correctly decide when this could actually become possible.

<p>AI failed to beat humans in 10 Math problems that expert mathematicians had solved in the past. Four systems entered a test, which was assessed by 30 analysts, and none of them was able to solve all 10 problems. The finding shows humans are still better than AI on new Math problems.</p>

Four AI models fail Math test they could not cheat on, humans score a perfect 10

<p>Families of Russian soldiers who died in combat during the Ukraine war are getting AI-generated photos and videos made of them, triggering a debate over the ethical complexities. These include fictional billboards, based on the idea of these soldiers returning home after the war is over. A popular blogger named Katya Jin recently posted a 15-second AI video showing a Russian soldier returning home to his family. Even Jin's own husband was fighting in the "special operation" and has been missing, and the AI video is apparently based on her. These photos and videos are being posted by the relatives of soldiers who have either died or have disappeared, and have left Ukrainians appalled, the BBC reported. These photos and videos mostly show the dead soldiers as heroes, with angel wings, leaving for paradise. While they could be a source of comfort for grieving families, some of whom also keep deepfakes featuring the deceased at funerals, most Ukrainians are appalled to see such content online.</p><p>&nbsp;</p><p>This "digital afterlife" industry has turned into a profitable venture for people making them. Content creators are charging anywhere from 200 roubles ($2.77) to 10,000 roubles ($138) per video. The bigger creators can earn up to 200,000 roubles ($2,800) a month, almost double the average Russian wage. This has led to criticism that they are capitalising on raw grief. Katya Jin used her own story to reach out to others, who are now ordering similar photos and videos.</p><h2><strong>Russia avoids talking about human loss in Ukraine war&nbsp;</strong></h2><p>A majority of these photos focus on soldiers who have died on the frontline. A total of 225,000 Russian soldiers are believed to have died in the four years of the conflict, and the number could be higher. The administration has received flak for the massive human loss, yet has avoided the topic publicly. However, the videos focus on patriotism and use AI to paint the soldiers as heroes who died for their country. The photos and videos depict them often as a man in uniform, embracing his family and descending up the stairs to heaven.&nbsp;</p>

AI photos and videos showing deceased Russian soldiers with wings, embracing their family and descending the stairs to heaven have gone viral. Some also show the ones still on the frontline reuniting with their loved ones.   

Families of Russian soldiers who died in combat during the Ukraine war are getting AI-generated photos and videos made of them, triggering a debate over the ethical complexities. These include fictional billboards, based on the idea of these soldiers returning home after the war is over. A popular blogger named Katya Jin recently posted a 15-second AI video showing a Russian soldier returning home to his family. Even Jin's own husband was fighting in the "special operation" and has been missing, and the AI video is apparently based on her. These photos and videos are being posted by the relatives of soldiers who have either died or have disappeared, and have left Ukrainians appalled, the BBC reported. These photos and videos mostly show the dead soldiers as heroes, with angel wings, leaving for paradise. While they could be a source of comfort for grieving families, some of whom also keep deepfakes featuring the deceased at funerals, most Ukrainians are appalled to see such content online. This "digital afterlife" industry has turned into a profitable venture for people making them. Content creators are charging anywhere from 200 roubles ($2.77) to 10,000 roubles ($138) per video. The bigger creators can earn up to 200,000 roubles ($2,800) a month, almost double the average Russian wage. This has led to criticism that they are capitalising on raw grief. Katya Jin used her own story to reach out to others, who are now ordering similar photos and videos. Russia avoids talking about human loss in Ukraine war A majority of these photos focus on soldiers who have died on the frontline. A total of 225,000 Russian soldiers are believed to have died in the four years of the conflict, and the number could be higher. The administration has received flak for the massive human loss, yet has avoided the topic publicly. However, the videos focus on patriotism and use AI to paint the soldiers as heroes who died for their country. The photos and videos depict them often as a man in uniform, embracing his family and descending up the stairs to heaven.

<p>AI photos and videos showing deceased Russian soldiers with wings, embracing their family and descending the stairs to heaven have gone viral. Some also show the ones still on the frontline reuniting with their loved ones. &nbsp;&nbsp;</p>

Russians spending thousands of roubles on AI photos and videos to resurrect dead soldiers

<p>As artificial intelligence drives a global race to build bigger and more powerful data centres, questions are growing about the resources needed to run them. Now, Amazon has released a key figure that it had never publicly disclosed before: its global data-centre operations used about 2.5 billion gallons, or more than 9 billion litres, of water in 2025.</p><p>&nbsp;</p><p>The announcement comes at a time when governments, researchers and local communities are paying closer attention to the environmental footprint of data centres, particularly those supporting AI systems.</p><p>&nbsp;</p><p>Amazon said water consumption at data centres it directly owns and operates actually fell by 2 per cent compared with 2024 levels, despite the company expanding its data-centre footprint. The disclosure was made through a company blog post as scrutiny of AI infrastructure continues to intensify.</p><p>&nbsp;</p><p>Data centres require significant amounts of electricity and cooling. As AI models become larger and more complex, the facilities powering them are attracting increased attention over their impact on water supplies and energy networks. Public concern is also growing. A recent Reuters poll found that while one in three Americans support data-centre construction, only 14 per cent said they would be comfortable having one built near where they live.</p><figure class="image"><img src="https://assets.aboutamazon.com/dims4/default/cf9712d/2147483647/strip/true/crop/1600x900+0+0/resize/1320x743!/quality/90/?url=https%3A%2F%2Famazon-blogs-brightspot.s3.amazonaws.com%2Fe0%2F93%2F3dccab4646e5bedc23c3905c4781%2Fwater-efficiency-table4-amazon-news-lt-060826.jpeg" alt="Water efficiency comparison table showing Amazon/AWS improving from 0.25 liters per kilowatt-hour in 2021 to 0.12 in 2025, outperforming Microsoft, Google, and Meta"></figure>

<h2><strong>Amazon says it uses less water than rivals</strong></h2><p>Amazon also used the disclosure to highlight what it says is its efficiency advantage. According to figures shared by the company, its data centres used 0.12 litres of water per kilowatt-hour of electricity in 2025. Amazon says this was lower than comparable figures published by rivals including Microsoft, Google and Meta. However, comparisons come with important caveats.</p><p>Google's figures cited by Amazon focus on Gemini AI-related data centres, which generally require more cooling because of their use of powerful AI processors. Amazon's reported water use also does not include water used during data-centre construction or water consumed by power plants generating electricity for its facilities.</p><p>&nbsp;</p><p>Amazon currently operates around 924 data centres globally, according to industry reports.</p><p>The wider debate over AI and water</p><p>The discussion extends well beyond Amazon.</p><p>Researchers at the United Nations University Institute for Water, Environment and Health estimate that the global water footprint of data centres could reach 9.3 trillion litres annually by 2030. According to the institute, that volume would be enough to meet the yearly domestic water needs of all 1.3 billion people living in sub-Saharan Africa.</p><figure class="image"><img src="https://assets.aboutamazon.com/dims4/default/f4ffae1/2147483647/strip/true/crop/1600x900+0+0/resize/1320x743!/quality/90/?url=https%3A%2F%2Famazon-blogs-brightspot.s3.amazonaws.com%2Fdf%2F2c%2F0d95222540f2b2dc51e7f7dc4cd3%2Fwater-efficiency-inline-graphic-numbers-amazon-news-lt-060926.jpeg" alt="Amazon water efficiency statistics: 10% cooling hours with evaporative design, 5.8 billion gallons returned annually, 130 data centers using reclaimed water"></figure><h2><strong>At the same time, transparency around data-centre water use is increasing.</strong></h2><p>In Oregon, legal action led to the release of records detailing water consumption linked to Google's facilities. Meanwhile, Utah recently passed legislation requiring certain new data-centre projects to disclose annual water usage.</p><p>Amazon's plan to become 'water positive' Amazon says it wants to return more water to communities than it consumes.</p><p>The company aims to become water positive by 2030 and says it has already achieved about 75 per cent of that goal.</p><p>Part of the strategy involves using reclaimed and treated wastewater rather than drinking water. Amazon says 26 sites now operate entirely on reclaimed water, while agreements have been signed for another 130 facilities worldwide.</p><p>The company also argues that its cooling systems reduce water demand. According to Amazon, its data centres rely on outside air cooling roughly 90 per cent of the time and only use water-based evaporative cooling during the hottest periods of the year.</p><p>&nbsp;</p><p>The disclosure highlights a growing challenge facing the technology industry. AI services are becoming more powerful and more widely used, but the infrastructure supporting them requires enormous amounts of electricity, computing power and water. As governments push for greater transparency and communities demand more information, companies such as Amazon, Google and Microsoft may face increasing pressure to show not only how much AI can achieve, but also what resources it consumes.</p>

Amazon has disclosed for the first time that its global data centres used more than 9 billion litres of water in 2025. The announcement comes as AI-driven data-centre growth faces increasing scrutiny over its impact on water resources, energy use and local communities.

As artificial intelligence drives a global race to build bigger and more powerful data centres, questions are growing about the resources needed to run them. Now, Amazon has released a key figure that it had never publicly disclosed before: its global data-centre operations used about 2.5 billion gallons, or more than 9 billion litres, of water in 2025. The announcement comes at a time when governments, researchers and local communities are paying closer attention to the environmental footprint of data centres, particularly those supporting AI systems. Amazon said water consumption at data centres it directly owns and operates actually fell by 2 per cent compared with 2024 levels, despite the company expanding its data-centre footprint. The disclosure was made through a company blog post as scrutiny of AI infrastructure continues to intensify. Data centres require significant amounts of electricity and cooling. As AI models become larger and more complex, the facilities powering them are attracting increased attention over their impact on water supplies and energy networks. Public concern is also growing. A recent Reuters poll found that while one in three Americans support data-centre construction, only 14 per cent said they would be comfortable having one built near where they live.

<p>Amazon has disclosed for the first time that its global data centres used more than 9 billion litres of water in 2025. The announcement comes as AI-driven data-centre growth faces increasing scrutiny over its impact on water resources, energy use and local communities.</p>

'9 billion litres': Here's how much water Amazon's data centres used in 2025

<p>When people think about artificial intelligence, they often imagine software engineers, data scientists and computer programmers. But according to Google and other technology giants, the future of AI depends just as much on plumbers, electricians, welders and construction workers.</p><p>Google has announced a $50 million investment in skilled-trades training programmes across the United States. The initiative aims to prepare workers for jobs that are becoming increasingly important as companies race to build the infrastructure needed to support artificial intelligence.</p><p>The announcement comes just days after Meta unveiled a $250 million programme focused on training workers for data centre construction jobs. Together, the investments highlight a growing reality: the AI boom is creating demand far beyond Silicon Valley.</p><p>&nbsp;</p>

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">America’s digital economy relies on our physical infrastructure and the electricians, pipefitters, welders, manufacturing workers &amp; more who build and maintain it. <br><br>Today, we’re making an additional <a href="https://x.com/Googleorg?ref_src=twsrc%5Etfw">@googleorg</a> commitment to help 300K American workers prepare for these in-demand…</p>&mdash; Sundar Pichai (@sundarpichai) <a href="https://x.com/sundarpichai/status/2065110608090161375?ref_src=twsrc%5Etfw">June 11, 2026</a></blockquote> <script async src="https://platform.x.com/widgets.js" charset="utf-8"></script>

<h2><strong>Why AI companies suddenly need blue-collar workers</strong></h2><p>The rapid growth of artificial intelligence has triggered a global race to build data centres, power facilities and digital infrastructure. Every advanced AI model requires enormous computing power. Behind every chatbot, AI assistant or image generator sits a vast network of servers operating inside large data centres.</p><p>Building those facilities requires much more than software. Electricians are needed to install and maintain power systems. Plumbers and pipe fitters help build cooling systems that prevent servers from overheating. Welders and construction workers are required to construct the buildings themselves.</p><p>Without these workers, many AI projects simply cannot move forward.</p><p>As Google CEO Sundar Pichai noted while announcing the programme, skilled trades will play a critical role in building the infrastructure needed for the next generation of technology.</p><p>A worker shortage is becoming a major problem. The construction industry is already struggling to find enough workers.</p><p>According to the Associated Builders and Contractors (ABC), the United States construction sector will need around 349,000 additional workers this year to meet demand. That demand is being fuelled by several trends at once, including AI data centre construction, energy projects, semiconductor manufacturing and broader infrastructure development. Industry experts say the shortage has become one of the biggest obstacles to future growth.</p><p>"The constraint on growth isn't hiring more engineers. It's building physical infrastructure," Tulane University business professor Rob Lalka said while discussing the issue. His comments highlight an important shift in the AI conversation. While much attention remains focused on software development, the physical systems that power AI are becoming equally important.</p><h2><strong>What jobs will Google's programme support?</strong></h2><p>Google says the funding will support training programmes focused on occupations critical to AI and energy infrastructure. These include electricians, plumbers, pipe fitters, welders, construction workers and other skilled trades. Some partnerships with training organisations are already underway, according to the company. The goal is to help create a larger workforce capable of supporting the growing number of data centres and energy projects being developed across the country.</p><p>For many workers, these jobs offer an alternative pathway into the technology economy without requiring a university degree in computer science.</p><h2><strong>Why data centres are driving demand</strong></h2><p>The AI industry is currently investing hundreds of billions of dollars in new infrastructure. Companies including Google, Meta, Microsoft, Amazon, Oracle and OpenAI are expanding data centre capacity to support increasingly powerful AI systems. Modern AI models require huge amounts of computing power for both training and daily operations. That means companies must continue building new facilities equipped with advanced chips, networking equipment and cooling systems.</p><p>Each new data centre creates demand for skilled trades workers long before software engineers begin using the infrastructure. This is one reason why technology companies are becoming more directly involved in workforce training programmes.</p><h2><strong>Meta, Microsoft and Oracle are making similar moves</strong></h2><p>Google is not alone. Meta recently announced a $250 million programme aimed at preparing workers for data centre construction projects. Earlier this year, Microsoft and Oracle also expanded workforce initiatives focused on supporting AI infrastructure growth. The growing number of programmes suggests that labour shortages are becoming a concern across the technology sector. Companies appear increasingly aware that future AI growth may depend not only on breakthroughs in software but also on having enough people available to build and maintain physical infrastructure.</p><h2><strong>The AI boom is creating new opportunities</strong></h2><p>The rise of artificial intelligence is often discussed in terms of automation and job displacement. However, the latest investments from Google and other technology companies highlight another side of the story. Building AI systems is creating demand for thousands of skilled workers across construction, energy and infrastructure sectors. For electricians, plumbers, welders and construction professionals, the AI boom may provide new career opportunities that did not exist just a few years ago. As technology companies continue investing in data centres and power infrastructure, these workers could become some of the most important people behind the next generation of artificial intelligence. The message from Google and its rivals is becoming increasingly clear: AI may run on software, but it is built by people working far beyond the walls of Silicon Valley.</p>

Google is investing $50 million to train electricians, plumbers, welders and construction workers as AI fuels a massive data centre boom. The future of artificial intelligence depends on skilled workers building the infrastructure behind it.

When people think about artificial intelligence, they often imagine software engineers, data scientists and computer programmers. But according to Google and other technology giants, the future of AI depends just as much on plumbers, electricians, welders and construction workers. Google has announced a $50 million investment in skilled-trades training programmes across the United States. The initiative aims to prepare workers for jobs that are becoming increasingly important as companies race to build the infrastructure needed to support artificial intelligence. The announcement comes just days after Meta unveiled a $250 million programme focused on training workers for data centre construction jobs. Together, the investments highlight a growing reality: the AI boom is creating demand far beyond Silicon Valley.

<p>Google is investing $50 million to train electricians, plumbers, welders and construction workers as AI fuels a massive data centre boom. The future of artificial intelligence depends on skilled workers building the infrastructure behind it.</p>

Google's latest AI investment isn't in software, it's in construction workers: Here's why it's investing in blue-collar workers

A new assessment of 100 AI agents reveals serious security weaknesses, showing that most can be manipulated through prompt injection attacks. The study highlights how hidden instructions inside emails or documents can mislead AI systems that have access to sensitive tools and data. Experts warn that as AI agents take on more real-world tasks, security gaps could become a major risk for organisations.


AI Fails Simple Test of Malicious Malware 

Artificial intelligence is increasingly being used to speed up daily clinical workloads, helping hospitals manage tasks such as patient data processing, diagnostics support, and administrative duties. However, despite rapid adoption, many healthcare systems are still struggling with a lack of proper AI training for medical staff. Experts warn that without structured training and guidelines, the full potential of AI in healthcare may not be safely or effectively realized.


AI is Used for Speeding Up Daily Clinical Workloads | AI Training Still Lacking in Hospitals

<p>When Anthropic unveiled the Mythos class of AI models, the internet asked one question louder than all others: what happens if you ask it to build a nuclear bomb? It is a fair question. Claude Fable 5, the public face of the Mythos class, released June 9, 2026, is described by Anthropic itself as the most capable AI model it has ever deployed. It scored 80.3% on SWE-Bench Pro, crossed the 90% threshold on analytical benchmarks, and outperformed every prior model on complex, long-horizon scientific tasks. So can it synthesise uranium? The short answer is no. The more interesting answer explains why.</p><p>&nbsp;</p><p>Anthropic has built what it calls a classifier-based fallback system directly into Fable 5. A separate layer of AI systems continuously monitors incoming prompts, looking for signals associated with CBRN misuse, chemical, biological, radiological, and nuclear domains. If a request crosses that threshold, the main Mythos-class model never sees it. The query is automatically rerouted to Claude Opus 4.8, which operates under tighter generation rules and returns a safe, non-actionable response. The underlying model's superior reasoning capability is simply never engaged. You are not talking to Mythos on those questions. You are talking to a gatekeeper.</p><p>&nbsp;</p><p>This architecture is not accidental. It is the direct output of Anthropic's Responsible Scaling Policy, which defines AI Safety Levels based on a model's potential to provide what the company calls ‘meaningful uplift’ in weapons of mass destruction. Claude Mythos 5, the unrestricted version of the same model, available only to a small group of vetted government and critical infrastructure partners, was evaluated against ASL-4 thresholds before any deployment decision was made. Anthropic concluded Mythos does not cross the ASL-4 line today, but it is the first model in history where the company judged the question close enough to require that formal evaluation at all.</p><p>&nbsp;</p><p>The nuclear chemistry guardrail is among the most absolute in the system. A user asking how to synthesize ricin, a biological toxin, triggers the fallback. A user asking about uranium enrichment centrifuge geometry triggers it. A user probing the critical mass calculations for weapons-grade plutonium triggers it. These are not edge cases that slip through. They are the primary design targets of the classifier layer.</p><p>&nbsp;</p><p>Anthropic has also been transparent about the limits of its own confidence. In a statement accompanying the Fable 5 launch, the company acknowledged that no classifier system is perfect, that jailbreak research is ongoing, and that a model of this capability class operating near weapons-relevant knowledge domains requires continuous red-teaming. The guardrails are real. They are also under constant adversarial pressure.</p><p>&nbsp;</p><p>What Claude Mythos represents is not a new danger in the nuclear space, that concern predates AI by decades. What it represents is a new class of machine intelligence powerful enough that the question had to be asked seriously, tested rigorously, and answered in public. Anthropic answered it. The Mythos species is not building bombs. The architecture ensures it.</p>

Claude Mythos and Fable 5 are Anthropic's most powerful models ever, but classifier-based CBRN guardrails ensure they cannot provide nuclear or chemical weapons uplift. Here's how the system works.

When Anthropic unveiled the Mythos class of AI models, the internet asked one question louder than all others: what happens if you ask it to build a nuclear bomb? It is a fair question. Claude Fable 5, the public face of the Mythos class, released June 9, 2026, is described by Anthropic itself as the most capable AI model it has ever deployed. It scored 80.3% on SWE-Bench Pro, crossed the 90% threshold on analytical benchmarks, and outperformed every prior model on complex, long-horizon scientific tasks. So can it synthesise uranium? The short answer is no. The more interesting answer explains why. Anthropic has built what it calls a classifier-based fallback system directly into Fable 5. A separate layer of AI systems continuously monitors incoming prompts, looking for signals associated with CBRN misuse, chemical, biological, radiological, and nuclear domains. If a request crosses that threshold, the main Mythos-class model never sees it. The query is automatically rerouted to Claude Opus 4.8, which operates under tighter generation rules and returns a safe, non-actionable response. The underlying model's superior reasoning capability is simply never engaged. You are not talking to Mythos on those questions. You are talking to a gatekeeper. This architecture is not accidental. It is the direct output of Anthropic's Responsible Scaling Policy, which defines AI Safety Levels based on a model's potential to provide what the company calls ‘meaningful uplift’ in weapons of mass destruction. Claude Mythos 5, the unrestricted version of the same model, available only to a small group of vetted government and critical infrastructure partners, was evaluated against ASL-4 thresholds before any deployment decision was made. Anthropic concluded Mythos does not cross the ASL-4 line today, but it is the first model in history where the company judged the question close enough to require that formal evaluation at all. The nuclear chemistry guardrail is among the most absolute in the system. A user asking how to synthesize ricin, a biological toxin, triggers the fallback. A user asking about uranium enrichment centrifuge geometry triggers it. A user probing the critical mass calculations for weapons-grade plutonium triggers it. These are not edge cases that slip through. They are the primary design targets of the classifier layer. Anthropic has also been transparent about the limits of its own confidence. In a statement accompanying the Fable 5 launch, the company acknowledged that no classifier system is perfect, that jailbreak research is ongoing, and that a model of this capability class operating near weapons-relevant knowledge domains requires continuous red-teaming. The guardrails are real. They are also under constant adversarial pressure. What Claude Mythos represents is not a new danger in the nuclear space, that concern predates AI by decades. What it represents is a new class of machine intelligence powerful enough that the question had to be asked seriously, tested rigorously, and answered in public. Anthropic answered it. The Mythos species is not building bombs. The architecture ensures it.

<p>Claude Mythos and its public version Fable 5 are Anthropic's most powerful AI models ever — but they are architecturally blocked from providing meaningful assistance with nuclear, chemical, biological, or radiological weapons. Here's how those guardrails actually work.</p>

Can Claude Mythos build a nuclear bomb? Inside Anthropic's hard chemistry guardrails