<p>Amid rising fears of the threat posed by AI, OpenAI has revealed that an autonomous AI agent powered by its advanced models went rogue during a security test and carried out a cyberattack on AI platform Hugging Face. The incident has raised fresh concerns over the risks involved in using advanced artificial intelligence systems.</p><h2><strong>What happened during the incident?</strong></h2><p>In a blog post on Tuesday (Jul 21), OpenAI explained an “unprecedented cyber incident” &nbsp;involving advanced AI-driven hacking capabilities. The company said that the incident occurred during an internal cybersecurity evaluation designed to test the limits of its AI models’ cyber abilities. The company said the models were placed in a highly isolated environment but managed to exploit a zero-day vulnerability in a package registry cache proxy to gain internet access.</p><p>The AI agents then identified and chained multiple vulnerabilities across OpenAI’s research systems and Hugging Face’s infrastructure while attempting to complete their assigned task. OpenAI said the models used stolen credentials and other attack methods to access restricted information from Hugging Face’s production environment. The company said the activity was detected by security teams, while Hugging Face’s security systems also identified and contained the breach.</p><p><strong>Also read |</strong><a href="https://www.wionews.com/world/missed-the-mark-meta-removes-instagram-ai-image-feature-after-backlash-1783737556704"><strong> ‘Missed the mark’: Meta removes Instagram AI image feature after backlash over privacy</strong></a></p><h2><strong>What did OpenAI say?</strong></h2><p>Following the incident, OpenAI said it was strengthening its safeguards and improving security measures. It added that it was tightening infrastructure controls, patching a disclosed zero-day vulnerability, and enhancing AI safety measures. Both companies said that they were investigating the incident.</p><p>“The primary lesson from this incident is that model security and safety must keep pace with rapidly advancing capabilities. We are strengthening the containment, monitoring, access controls, and evaluation practices used during model development,” OpenAI said.</p><p>Hugging Face, a popular platform that hosts open-source AI models and datasets, had earlier disclosed that it suffered a sophisticated cyberattack unlike previous incidents. The company said the attack appeared to have been conducted entirely by an autonomous AI agent.</p><p><strong>Also read | </strong><a href="https://www.wionews.com/india-news/india-orders-meta-to-remove-child-sexual-abuse-ads-from-instagram-demands-explaination-in-7-days-1783235163711"><strong>India orders Meta to remove Instagram ads promoting child sexual abuse, demands explanation in 7 days: Report</strong></a></p><p>In a statement, Hugging Face Co-founder and CEO Clem Delangue said, “We’re grateful for the collaboration with OpenAI on this and other topics. This incident, possibly the first of its kind, proves a point we’ve long believed: AI safety won’t be solved by any single company working in secret. It will be solved in the open, collaboratively, with broad access to AI for every defender, everywhere.”</p><h2><strong>Concerns over AI</strong></h2><p>The disclosure has intensified concerns among cybersecurity experts and policymakers about the ability of frontier AI models to operate independently and potentially cause harm.</p><p>US Representative Greg Casar called the incident alarming and urged stronger safeguards, including mandatory independent safety testing, reporting of AI-related security breaches and global cooperation on AI regulation.</p><p>Cybersecurity experts warned that the incident could represent a preview of future threats. Katie Moussouris, CEO of Luta Security, compared advanced AI models to highly adaptive systems capable of finding ways around restrictions.</p><p>Matt Suiche, an engineer at AI cybersecurity firm Tolmo, said the incident showed that frontier AI models are rapidly approaching the capabilities of sophisticated attackers. He added that similar AI-driven cyber risks are no longer limited to top research labs and can emerge from widely available technologies.</p><p><strong>Watch | The Internet’s Darkest Secret: Inside the global network exploiting children online</strong></p><p><a href="https://www.wionews.com/videos/the-internet-s-darkest-secret-inside-the-global-network-exploiting-children-online-1784689947807/embed">https://www.wionews.com/videos/the-internet-s-darkest-secret-inside-the-global-network-exploiting-children-online-1784689947807/embed</a></p><p>&nbsp;</p>

OpenAI said an AI agent went rogue during a security test, escaped containment and hacked Hugging Face’s infrastructure, raising concerns over the cybersecurity risks posed by advanced autonomous AI systems.

Amid rising fears of the threat posed by AI, OpenAI has revealed that an autonomous AI agent powered by its advanced models went rogue during a security test and carried out a cyberattack on AI platform Hugging Face. The incident has raised fresh concerns over the risks involved in using advanced artificial intelligence systems. What happened during the incident? In a blog post on Tuesday (Jul 21), OpenAI explained an “unprecedented cyber incident” involving advanced AI-driven hacking capabilities. The company said that the incident occurred during an internal cybersecurity evaluation designed to test the limits of its AI models’ cyber abilities. The company said the models were placed in a highly isolated environment but managed to exploit a zero-day vulnerability in a package registry cache proxy to gain internet access. The AI agents then identified and chained multiple vulnerabilities across OpenAI’s research systems and Hugging Face’s infrastructure while attempting to complete their assigned task. OpenAI said the models used stolen credentials and other attack methods to access restricted information from Hugging Face’s production environment. The company said the activity was detected by security teams, while Hugging Face’s security systems also identified and contained the breach. Also read | ‘Missed the mark’: Meta removes Instagram AI image feature after backlash over privacy What did OpenAI say? Following the incident, OpenAI said it was strengthening its safeguards and improving security measures. It added that it was tightening infrastructure controls, patching a disclosed zero-day vulnerability, and enhancing AI safety measures. Both companies said that they were investigating the incident. “The primary lesson from this incident is that model security and safety must keep pace with rapidly advancing capabilities. We are strengthening the containment, monitoring, access controls, and evaluation practices used during model development,” OpenAI said. Hugging Face, a popular platform that hosts open-source AI models and datasets, had earlier disclosed that it suffered a sophisticated cyberattack unlike previous incidents. The company said the attack appeared to have been conducted entirely by an autonomous AI agent. Also read | India orders Meta to remove Instagram ads promoting child sexual abuse, demands explanation in 7 days: Report In a statement, Hugging Face Co-founder and CEO Clem Delangue said, “We’re grateful for the collaboration with OpenAI on this and other topics. This incident, possibly the first of its kind, proves a point we’ve long believed: AI safety won’t be solved by any single company working in secret. It will be solved in the open, collaboratively, with broad access to AI for every defender, everywhere.” Concerns over AI The disclosure has intensified concerns among cybersecurity experts and policymakers about the ability of frontier AI models to operate independently and potentially cause harm. US Representative Greg Casar called the incident alarming and urged stronger safeguards, including mandatory independent safety testing, reporting of AI-related security breaches and global cooperation on AI regulation. Cybersecurity experts warned that the incident could represent a preview of future threats. Katie Moussouris, CEO of Luta Security, compared advanced AI models to highly adaptive systems capable of finding ways around restrictions. Matt Suiche, an engineer at AI cybersecurity firm Tolmo, said the incident showed that frontier AI models are rapidly approaching the capabilities of sophisticated attackers. He added that similar AI-driven cyber risks are no longer limited to top research labs and can emerge from widely available technologies. Watch | The Internet’s Darkest Secret: Inside the global network exploiting children online https://www.wionews.com/videos/the-internet-s-darkest-secret-inside-the-global-network-exploiting-children-online-1784689947807/embed

<p>OpenAI said an AI agent went rogue during a security test, escaped containment and hacked Hugging Face’s infrastructure, raising concerns over the cybersecurity risks posed by advanced autonomous AI systems.</p>

OpenAI’s AI agent gone rogue? Company reveals chilling hack at startup

<p>Google's race to stay at the front of artificial intelligence has reportedly hit another obstacle. According to a Bloomberg report, the company has delayed the release of Gemini 3.5 Pro, its next flagship AI model, after it failed to meet internal performance targets. The biggest concern was said to be coding an area where rivals OpenAI and Anthropic have recently gained significant momentum. The model was expected to launch in June, shortly after CEO Sundar Pichai introduced it during Google I/O 2026, saying it was already being used internally. Months later, however, Gemini 3.5 Pro is still undergoing testing.</p><p>The report unsettled investors. Shares of Google's parent company, Alphabet, fell nearly 3 per cent during trading and were down about 4 per cent at one stage.</p><h2><strong>Coding has become the new AI battleground</strong></h2><p>The delay highlights how important coding has become in the AI industry. Large language models are no longer judged only by how well they answer questions or generate text. Increasingly, companies want AI systems that can write, review, debug and improve software with minimal human input. According to Bloomberg, Google retrained Gemini with updated data last month to improve its coding ability, but engineers were reportedly disappointed with the results. Meanwhile, OpenAI's GPT-5.6 and Anthropic's Claude Fable 5 have raised expectations across the industry, particularly for enterprise software development.</p><h2><strong>Internal challenges may also be slowing Google</strong></h2><p>The report suggests Google's size has become one of its biggest challenges. Unlike AI startups that focus on a single product, Google must coordinate across Search, DeepMind, Cloud, Android, YouTube and other divisions before launching major AI models. One former employee reportedly described the process as trying to "boil an ocean," pointing to overlapping teams, changing priorities and internal competition for computing resources. Google has also tightened oversight around AI development as advanced models face growing scrutiny from governments over safety, national security and misuse.</p><h2><strong>Google says testing is still underway</strong></h2><p>Google has not confirmed a launch date for Gemini 3.5 Pro. A company spokesperson told Bloomberg that Google is "shipping quickly across a wide range of models while keeping them highly cost-effective" and confirmed that Gemini 3.5 Pro, an upgraded Flash model and other AI systems are currently being tested with partners. The spokesperson also said Google remains engaged with the US government as oversight of frontier AI models increases.</p><h2><strong>Is Google really falling behind?</strong></h2><p>Despite the reported delay, industry observers say Google remains one of the strongest AI players. Its biggest advantage is deep integration across products used by billions of people, including Search, Gmail, Maps, Android, Chrome and Workspace. Gemini also benefits from Google's search index, giving it access to one of the world's largest information ecosystems.</p><p>Google recently revealed that around 75 per cent of its internal code is now generated with AI assistance before being reviewed by engineers. The company has also consolidated several coding initiatives under its Antigravity platform to accelerate software development. However, with OpenAI and Anthropic releasing increasingly capable models, expectations for Gemini 3.5 Pro have only grown. The reported delay suggests Google is choosing to refine the model rather than launch a product that falls short of its own standards a decision that could prove important as the AI race enters an even more competitive phase.</p>

Google has reportedly delayed the launch of Gemini 3.5 Pro after the AI model failed to meet internal expectations, particularly in coding.

Google's race to stay at the front of artificial intelligence has reportedly hit another obstacle. According to a Bloomberg report, the company has delayed the release of Gemini 3.5 Pro, its next flagship AI model, after it failed to meet internal performance targets. The biggest concern was said to be coding an area where rivals OpenAI and Anthropic have recently gained significant momentum. The model was expected to launch in June, shortly after CEO Sundar Pichai introduced it during Google I/O 2026, saying it was already being used internally. Months later, however, Gemini 3.5 Pro is still undergoing testing. The report unsettled investors. Shares of Google's parent company, Alphabet, fell nearly 3 per cent during trading and were down about 4 per cent at one stage. Coding has become the new AI battleground The delay highlights how important coding has become in the AI industry. Large language models are no longer judged only by how well they answer questions or generate text. Increasingly, companies want AI systems that can write, review, debug and improve software with minimal human input. According to Bloomberg, Google retrained Gemini with updated data last month to improve its coding ability, but engineers were reportedly disappointed with the results. Meanwhile, OpenAI's GPT-5.6 and Anthropic's Claude Fable 5 have raised expectations across the industry, particularly for enterprise software development. Internal challenges may also be slowing Google The report suggests Google's size has become one of its biggest challenges. Unlike AI startups that focus on a single product, Google must coordinate across Search, DeepMind, Cloud, Android, YouTube and other divisions before launching major AI models. One former employee reportedly described the process as trying to "boil an ocean," pointing to overlapping teams, changing priorities and internal competition for computing resources. Google has also tightened oversight around AI development as advanced models face growing scrutiny from governments over safety, national security and misuse. Google says testing is still underway Google has not confirmed a launch date for Gemini 3.5 Pro. A company spokesperson told Bloomberg that Google is "shipping quickly across a wide range of models while keeping them highly cost-effective" and confirmed that Gemini 3.5 Pro, an upgraded Flash model and other AI systems are currently being tested with partners. The spokesperson also said Google remains engaged with the US government as oversight of frontier AI models increases. Is Google really falling behind? Despite the reported delay, industry observers say Google remains one of the strongest AI players. Its biggest advantage is deep integration across products used by billions of people, including Search, Gmail, Maps, Android, Chrome and Workspace. Gemini also benefits from Google's search index, giving it access to one of the world's largest information ecosystems. Google recently revealed that around 75 per cent of its internal code is now generated with AI assistance before being reviewed by engineers. The company has also consolidated several coding initiatives under its Antigravity platform to accelerate software development. However, with OpenAI and Anthropic releasing increasingly capable models, expectations for Gemini 3.5 Pro have only grown. The reported delay suggests Google is choosing to refine the model rather than launch a product that falls short of its own standards a decision that could prove important as the AI race enters an even more competitive phase.

<p>Google has reportedly delayed the launch of Gemini 3.5 Pro after the AI model failed to meet internal expectations, particularly in coding. The setback comes as OpenAI and Anthropic continue releasing more advanced models, increasing pressure on Google in one of the industry's most competitive races.</p>

Google delays Gemini 3.5 Pro after AI model misses coding targets, report says

<p>The race to build more powerful artificial intelligence models is about to get even more competitive. Elon Musk has announced that Grok 4.5 will be released to the public tomorrow after completing private beta testing.</p><p>In a post on X, Musk said the launch follows "strong positive feedback" from users in the beta programme. He described Grok 4.5 as an "Opus-class" model, comparing it with Anthropic's Claude family of AI models, while claiming it is faster, more token-efficient and cheaper to run.</p><p>The announcement comes as AI companies continue to compete on speed, reasoning ability and operating costs rather than simply building larger models.</p>

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">Based on strong positive feedback from customers in our beta test program, <a href="https://x.com/SpaceXAI?ref_src=twsrc%5Etfw">@SpaceXAI</a> will make Grok 4.5 available to the public tomorrow.<br><br>It is an Opus-class model, but faster, more token-efficient and lower cost.</p>&mdash; Elon Musk (@elonmusk) <a href="https://x.com/elonmusk/status/2074740539874775163?ref_src=twsrc%5Etfw">July 8, 2026</a></blockquote> <script async src="https://platform.x.com/widgets.js" charset="utf-8"></script>

<h2><strong>What is new in Grok 4.5?</strong></h2><p><br>According to Musk, Grok 4.5 is built on SpaceXAI's new 1.5 trillion-parameter V9 foundation model, making it one of the company's most advanced AI systems. Earlier this month, Musk said the model had entered private beta testing at SpaceX and Tesla, where engineers tested its coding, reasoning and general-purpose capabilities before a wider rollout.</p><p>Although SpaceXAI has not released detailed benchmark results, Musk claims the new model delivers better performance while using fewer computing resources, making it cheaper to operate than previous versions.</p><h2><strong>Why is this launch important?</strong></h2><p>The AI industry is becoming increasingly competitive, Over the past few months, companies including OpenAI, Anthropic, Google and Meta have all released more capable AI models with improved reasoning and coding abilities.</p><p>Instead of focusing only on raw performance, developers are now trying to make models more efficient. Lower computing costs allow companies to serve more users while reducing the expensive infrastructure needed to run advanced AI systems. If Musk's claims prove accurate, Grok 4.5 could strengthen SpaceXAI's position in the rapidly growing AI market.</p><h2><strong>Competition continues to intensify</strong></h2><p>SpaceXAI, formerly known as xAI, has expanded rapidly since becoming part of SpaceX earlier this year. The company has continued updating its flagship Grok family, with Grok 4.3 launching in April and Grok 4.5 arriving just a few months later. Reuters also reported that SpaceXAI is working with AI coding platform Cursor as it expands its enterprise ambitions beyond consumer chatbots. With Grok 4.5 now ready for public release, the latest launch adds another major competitor to an AI market already dominated by ChatGPT, Claude and Google's Gemini. Whether Grok 4.5 can match Musk's promises will become clearer once users begin testing the model after its public rollout.</p>

Elon Musk has announced that Grok 4.5 will be released to the public tomorrow after private beta testing. The latest AI model is expected to offer faster performance, improved efficiency and lower operating costs as competition among leading AI companies continues to grow.

The race to build more powerful artificial intelligence models is about to get even more competitive. Elon Musk has announced that Grok 4.5 will be released to the public tomorrow after completing private beta testing. In a post on X, Musk said the launch follows "strong positive feedback" from users in the beta programme. He described Grok 4.5 as an "Opus-class" model, comparing it with Anthropic's Claude family of AI models, while claiming it is faster, more token-efficient and cheaper to run. The announcement comes as AI companies continue to compete on speed, reasoning ability and operating costs rather than simply building larger models.

<p>Elon Musk has announced that Grok 4.5 will be released to the public tomorrow after private beta testing. The latest AI model is expected to offer faster performance, improved efficiency and lower operating costs as competition among leading AI companies continues to grow.</p>

Elon Musk calls Grok 4.5 an 'Opus-class' AI model. Here's what that means

<p>The US government has approved limited access to Anthropic's advanced Mythos 5 artificial intelligence model after weeks of negotiations over national security concerns. The decision marks a major policy shift after Washington temporarily blocked access to Mythos 5 and Fable 5 earlier this month, citing fears that the models could be "jailbroken" and misused for cyberattacks. The latest approval allows Anthropic to restore Mythos 5 to a small group of trusted cybersecurity partners, while broader restrictions on Fable 5 remain unchanged.</p><h2><strong>Why the US changed its decision</strong></h2><p>According to a letter from US Commerce Secretary Howard Lutnick, Anthropic made significant progress in addressing risks linked to the advanced AI model. Anthropic said it is now working to restore access for approved cyber defenders and infrastructure providers as quickly as possible. The company added that discussions with the US government will continue as it seeks wider access for Mythos 5 and hopes to make Fable 5 generally available again. The company has not disclosed exactly what technical or security changes were made to satisfy government officials.</p><p>&nbsp;</p><h2><strong>Why Mythos 5 drew attention</strong></h2><p>Mythos 5 is Anthropic's most advanced cybersecurity-focused AI model. It was designed to help researchers identify software vulnerabilities and strengthen digital defences, but officials worried the same capabilities could be exploited by malicious actors. Before restrictions were imposed, around 200 organisations, including Apple, Google, Microsoft, Nvidia and Cisco, reportedly had access to the model through Anthropic's invite-only Project Glasswing programme.</p><p>&nbsp;</p><h2><strong>Bigger impact on the AI industry</strong></h2><p>The dispute became one of the biggest government interventions in the AI industry so far. It also influenced rival OpenAI, which recently introduced a limited preview of GPT-5.6 following discussions with US officials. Anthropic says it will continue working with regulators to create a long-term framework for releasing powerful AI systems safely. The outcome could shape how future frontier AI models are launched, balancing rapid innovation with national security concerns.</p>

The US has approved limited access to Anthropic's Mythos 5 AI model after the company addressed national security concerns. The decision eases restrictions imposed earlier this month, while Fable 5 remains restricted as discussions with regulators continue.


The US government has approved limited access to Anthropic's advanced Mythos 5 artificial intelligence model after weeks of negotiations over national security concerns. The decision marks a major policy shift after Washington temporarily blocked access to Mythos 5 and Fable 5 earlier this month, citing fears that the models could be "jailbroken" and misused for cyberattacks. The latest approval allows Anthropic to restore Mythos 5 to a small group of trusted cybersecurity partners, while broader restrictions on Fable 5 remain unchanged. Why the US changed its decision According to a letter from US Commerce Secretary Howard Lutnick, Anthropic made significant progress in addressing risks linked to the advanced AI model. Anthropic said it is now working to restore access for approved cyber defenders and infrastructure providers as quickly as possible. The company added that discussions with the US government will continue as it seeks wider access for Mythos 5 and hopes to make Fable 5 generally available again. The company has not disclosed exactly what technical or security changes were made to satisfy government officials. Why Mythos 5 drew attention Mythos 5 is Anthropic's most advanced cybersecurity-focused AI model. It was designed to help researchers identify software vulnerabilities and strengthen digital defences, but officials worried the same capabilities could be exploited by malicious actors. Before restrictions were imposed, around 200 organisations, including Apple, Google, Microsoft, Nvidia and Cisco, reportedly had access to the model through Anthropic's invite-only Project Glasswing programme. Bigger impact on the AI industry The dispute became one of the biggest government interventions in the AI industry so far. It also influenced rival OpenAI, which recently introduced a limited preview of GPT-5.6 following discussions with US officials. Anthropic says it will continue working with regulators to create a long-term framework for releasing powerful AI systems safely. The outcome could shape how future frontier AI models are launched, balancing rapid innovation with national security concerns.

<p>The US has approved limited access to Anthropic's Mythos 5 AI model after the company addressed national security concerns. The decision eases restrictions imposed earlier this month, while Fable 5 remains restricted as discussions with regulators continue.</p>

Anthropic's most powerful AI model gets partial US approval. Here's what changed

<p>The global artificial intelligence race may be entering a new phase. Just weeks after the US tightened restrictions on access to some of the world's most advanced AI models, a leading Chinese AI executive has suggested that China could soon develop comparable systems of its own.</p><p>&nbsp;</p><p>Tang Jie, founder of Chinese AI lab Zai, made the comments following the launch of GLM 5.2, the company's latest large language model. His remarks have attracted attention because they come amid growing competition between the United States and China over advanced AI capabilities.</p><p>&nbsp;</p><p>The comments also arrived after recent restrictions on access to Anthropic's most powerful AI models for foreign users, reportedly due to concerns that hostile actors could misuse them for cyber operations.</p><p>&nbsp;</p><h2><strong>A disagreement with Elon Musk</strong></h2><p>The discussion began on X when a user suggested that China could have a "Mythos-class" AI model by the end of 2026. Elon Musk responded by predicting that such a development was more likely around the first quarter of 2027.</p><p>Tang disagreed.</p><p>&nbsp;</p><p>"Won't take that long," he replied, signalling confidence that Chinese AI development may be progressing faster than many observers expect. The exchange quickly drew attention from AI researchers and technology analysts because it touched on one of the industry's most important questions: how quickly China can close the gap with leading American AI labs.</p><figure class="image"><img src="https://akm-img-a-in.tosshub.com/indiatoday/styles/medium_crop_simple/public/2026-06/zai.jpg?VersionId=UiqPmL7MixsEzppy.un7ng4UichBlaGH" alt="Zai"></figure>

<h2><strong>Why GLM 5.2 is generating attention</strong></h2><p>Much of the discussion centres on GLM 5.2 itself. According to benchmark data cited by supporters of the model, GLM 5.2 has become one of the highest-performing open-weight AI models currently available. Open-weight models make their learned parameters publicly available, allowing developers and researchers to run, study and modify them locally. GLM 5.2 reportedly achieved a score of 51 on the Artificial Analysis Intelligence Index. That places it ahead of several well-known models, including Google's Gemini 3.5 Flash and Anthropic's Claude Sonnet 4.6 on that benchmark. However, it still trails newer frontier systems such as Anthropic's Fable 5 and OpenAI's GPT 5.5.</p><h2><strong>Trained without Nvidia chips</strong></h2><p>One detail attracting particular interest is how the model was built. According to Zai, GLM 5.2 was trained entirely using Huawei Ascend chips rather than Nvidia hardware. That matters because US export restrictions have significantly limited China's access to some of Nvidia's most advanced AI processors. The success of a high-performing model trained on domestic hardware could be viewed as evidence that Chinese companies are finding alternative paths forward despite restrictions. Reports have also suggested the training costs were substantially lower than those associated with many leading Western AI models.</p><p>&nbsp;</p><p>The debate extends beyond benchmark scores. Advanced AI systems are increasingly viewed as strategic technologies with implications for cybersecurity, economic competitiveness and national security. American policymakers have argued that limiting access to cutting-edge AI systems can help reduce certain risks, including the misuse of powerful models for cyberattacks. However, if Chinese firms develop comparable capabilities independently, the long-term impact of those restrictions could become less clear. At the same time, leading US companies are unlikely to stand still.</p><p>Anthropic, OpenAI, Google and other major labs continue to release more capable systems at a rapid pace, meaning the benchmark for frontier AI is constantly moving higher.</p><h2><strong>The next chapter in the US-China AI competition</strong></h2><p>The exchange between Tang Jie and Elon Musk highlights how quickly the AI landscape is changing. Only a few years ago, the world's most advanced AI systems were concentrated among a handful of American companies. Today, Chinese firms are producing increasingly competitive models and attracting growing international attention. Whether China can develop a true frontier model within the timeline suggested by Tang remains uncertain. What is clear, however, is that the competition between the world's two largest AI ecosystems is accelerating, and the next major breakthrough may arrive sooner than many expected.</p><p>&nbsp;</p>

Chinese AI lab Zai says China could develop a frontier AI model comparable to today's most advanced systems sooner than expected. The comments follow the launch of GLM 5.2, a high-performing open-weight model trained entirely on Huawei chips, highlighting the growing US-China AI rivalry.

The global artificial intelligence race may be entering a new phase. Just weeks after the US tightened restrictions on access to some of the world's most advanced AI models, a leading Chinese AI executive has suggested that China could soon develop comparable systems of its own. Tang Jie, founder of Chinese AI lab Zai, made the comments following the launch of GLM 5.2, the company's latest large language model. His remarks have attracted attention because they come amid growing competition between the United States and China over advanced AI capabilities. The comments also arrived after recent restrictions on access to Anthropic's most powerful AI models for foreign users, reportedly due to concerns that hostile actors could misuse them for cyber operations. A disagreement with Elon Musk The discussion began on X when a user suggested that China could have a "Mythos-class" AI model by the end of 2026. Elon Musk responded by predicting that such a development was more likely around the first quarter of 2027. Tang disagreed. "Won't take that long," he replied, signalling confidence that Chinese AI development may be progressing faster than many observers expect. The exchange quickly drew attention from AI researchers and technology analysts because it touched on one of the industry's most important questions: how quickly China can close the gap with leading American AI labs.

<p>Chinese AI lab Zai says China could develop a frontier AI model comparable to today's most advanced systems sooner than expected. The comments follow the launch of GLM 5.2, a high-performing open-weight model trained entirely on Huawei chips, highlighting the growing US-China AI rivalry.</p>

China could have its own 'Mythos-class' AI sooner than expected, says Zai chief after GLM 5.2 launch

<p>Google has unveiled Gemma 4 12B, the latest version of its open AI model family, bringing advanced reasoning, vision and audio capabilities to devices that can run locally without depending entirely on cloud infrastructure.</p><p>According to Google, the new model delivers performance close to much larger AI systems while requiring significantly less memory. The company says Gemma 4 12B can run locally on devices equipped with just 16GB of VRAM, making advanced AI more accessible to developers, researchers and businesses. The launch highlights a growing trend across the AI industry: bringing powerful AI models directly to personal computers instead of relying solely on remote data centres.</p>

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">Today we’re introducing Gemma 4 12B — our latest open model that brings advanced agentic reasoning, vision and audio directly to your laptop.<br><br>It delivers performance nearing our larger Gemma models with a much smaller total memory footprint, while being small enough to run… <a href="https://t.co/M5xw74tEAW">pic.twitter.com/M5xw74tEAW</a></p>&mdash; Google (@Google) <a href="https://x.com/Google/status/2062203526588088452?ref_src=twsrc%5Etfw">June 3, 2026</a></blockquote> <script async src="https://platform.x.com/widgets.js" charset="utf-8"></script>

<h2><strong>What is Gemma 4 12B?</strong></h2><p>Gemma is Google's family of open AI models built using technology and research from its Gemini programme. The new Gemma 4 12B model contains 12 billion parameters and has been designed to handle multiple types of information, including text, images and audio. Unlike traditional AI systems that focus only on text, Gemma 4 12B can understand visual content, process audio inputs and perform advanced reasoning tasks. This makes it suitable for a wider range of applications, from software development and content creation to research and automation. Google says the model is available under the Apache 2.0 licence, allowing developers and organisations to use, modify and deploy it with relatively few restrictions.</p><h2><strong>Why is this launch important?</strong></h2><p>One of the biggest challenges in artificial intelligence today is the cost of running powerful models. Many advanced AI systems require expensive cloud infrastructure and high-performance servers. That can create barriers for smaller companies, independent developers and researchers. Google says Gemma 4 12B addresses this challenge by offering strong performance while remaining compact enough to run on consumer hardware. The company claims users can access advanced AI capabilities on their own machines without constantly sending requests to cloud servers.</p><p>This approach can provide several advantages:</p><p>Lower operating costs<br>Better privacy<br>Faster response times<br>Reduced internet dependence<br>Greater control over data</p><p>Industry experts increasingly see local AI as one of the next major shifts in computing. A new architecture removes traditional limitations</p><p>One of the most significant technical changes in Gemma 4 12B is its new unified architecture. Traditionally, multimodal AI systems use separate components known as encoders to process images, audio and text before combining the information. Google says Gemma 4 12B removes the need for separate multimodal encoders. Instead, the model processes different types of information through a unified architecture. According to the company, this helps improve efficiency while reducing memory requirements and computational overhead.</p><p>The result is a model that can deliver advanced multimodal capabilities while remaining small enough to run locally on modern hardware.</p><h2><strong>The growing race for local AI</strong></h2><p>Google's announcement comes as technology companies increasingly compete to bring AI closer to users. Microsoft recently introduced AI-focused hardware such as the Surface RTX Spark Dev Box and Surface Laptop Ultra. Apple has expanded on-device AI through Apple Intelligence, while companies including Perplexity and Nvidia are investing heavily in local AI computing.</p><p>The industry's focus is shifting because running every AI request through the cloud is becoming increasingly expensive. Local AI offers a way to reduce infrastructure costs while improving privacy and responsiveness. Gemma 4 12B positions Google as an important player in this emerging market.</p><h2><strong>Who could use Gemma 4 12B?</strong></h2><p>The model is expected to appeal to several groups.<br><br>Developers<br>Software developers can use the model to build AI-powered applications without relying entirely on external APIs.</p><p>Researchers&nbsp;<br>Researchers can experiment with advanced AI systems locally while maintaining greater control over their data.</p><p>Businesses<br>Companies working with sensitive information may prefer local AI processing for privacy and compliance reasons.</p><p>AI enthusiasts<br>Because the model is openly available, hobbyists and independent developers can also explore its capabilities without significant infrastructure investments.</p><h2><strong>What does this mean for the future of AI?</strong></h2><p>The release of Gemma 4 12B reflects a broader shift taking place across the industry. For years, AI development focused primarily on building larger and larger models in massive data centres. Increasingly, however, companies are trying to make powerful AI systems smaller, more efficient and easier to run on everyday devices. Google's latest model suggests that future AI progress may not only be about scale. Efficiency, accessibility and local deployment are becoming equally important.</p><p>If models like Gemma 4 12B continue to improve, users may soon be able to run advanced AI assistants, coding tools and productivity systems directly on their laptops without relying heavily on cloud services. For developers and businesses, that could open the door to a new generation of AI applications that are faster, more private and less expensive to operate.</p>

Google has unveiled Gemma 4 12B, the latest version of its open AI model family, bringing advanced reasoning, vision and audio capabilities to devices that can run locally without depending entirely on cloud infrastructure.


Google has unveiled Gemma 4 12B, the latest version of its open AI model family, bringing advanced reasoning, vision and audio capabilities to devices that can run locally without depending entirely on cloud infrastructure. According to Google, the new model delivers performance close to much larger AI systems while requiring significantly less memory. The company says Gemma 4 12B can run locally on devices equipped with just 16GB of VRAM, making advanced AI more accessible to developers, researchers and businesses. The launch highlights a growing trend across the AI industry: bringing powerful AI models directly to personal computers instead of relying solely on remote data centres.

<p>Google has unveiled Gemma 4 12B, the latest version of its open AI model family, bringing advanced reasoning, vision and audio capabilities to devices that can run locally without depending entirely on cloud infrastructure.</p>

Google launches Gemma 4 12B: This powerful AI model from Google can run on your laptop

<p>Meta has reportedly delayed the release of its new artificial intelligence model, codenamed Avocado, after the system failed to meet internal performance benchmarks.</p><p>According to a report by The New York Times, the company had initially planned to launch the model earlier this year. However, Meta has now pushed the rollout to at least May while engineers work to improve its performance.</p><p>The delay highlights the intense competition among major AI companies as they race to develop more powerful large language models.</p><h2><strong>Avocado falls behind rival AI models</strong></h2><p>Internal testing reportedly showed that Meta’s Avocado model struggled to match the latest systems from competitors.</p><p>According to the report, Avocado did not outperform Gemini 3, the AI model launched by Google last year. However, Avocado reportedly performed better than Gemini 2.5 and also improved upon Meta’s earlier models, such as Llama 4.</p><p>Industry analysts say these benchmark comparisons are critical because large language models compete directly on reasoning ability, coding performance and accuracy.</p><p>Meta has not publicly confirmed the exact performance scores from its internal tests.</p><h2><strong>Meta may temporarily licence Google’s AI models</strong></h2><p>The report also said that Meta executives have discussed the possibility of licensing Google’s Gemini models temporarily. This move could allow Meta to maintain the performance of its AI chatbot, Meta AI, while the company continues to develop Avocado. Meta AI is integrated across several major platforms including Facebook, Instagram, and WhatsApp. Licensing a rival model would help Meta avoid losing users to competing AI services. Such arrangements are becoming more common in the industry. For example, Apple recently announced plans to integrate Gemini into its Siri assistant.</p><h2><strong>Superintelligence Labs working on next-generation AI</strong></h2><p>Avocado is the first major model being developed by Meta’s new Superintelligence Labs (MSL) division. The unit is led by Alexandr Wang, whose startup Scale AI was acquired by Meta last year in a deal valued at about $14.5 billion.</p><p>Reports suggest Avocado is designed to focus on:</p><p>logical reasoning</p><p>software development tasks</p><p>agentic AI capabilities</p><p>Agentic systems are designed to perform multi-step tasks independently, rather than simply generating responses like traditional chatbots.</p><p>The division is also working on another AI model called Mango, which focuses on image generation.</p><h2><strong>Meta increases spending on AI infrastructure</strong></h2><p>Despite the delay, Meta continues to invest heavily in artificial intelligence development.</p><p>The company has announced plans to spend between $115 billion and $135 billion on AI infrastructure this year, including data centres, chips and computing systems needed to train advanced AI models.</p><p>Meta has also recently acquired the AI-focused social platform Moltbook, whose development team will join the Superintelligence Labs division.</p><p>The move reflects the broader strategy of expanding research into AI agents and next-generation models.</p><h2><strong>Growing competition in the AI industry</strong></h2><p>The delay of Avocado illustrates the growing pressure among technology companies competing in the AI race.</p><p>Major firms including OpenAI, Google, and Anthropic continue to release increasingly advanced models.</p><p>As these systems become more capable, companies are focusing not only on chatbots but also on AI agents capable of performing complex tasks.</p><p>For Meta, improving Avocado’s performance will be important to remain competitive in this rapidly evolving market.</p>

Meta has delayed the release of its new AI model Avocado after internal tests showed it failed to meet performance benchmarks against rival systems. According to reports, the model underperformed compared with Google’s Gemini 3, though it improved on earlier Meta models like Llama 4. 

Meta has reportedly delayed the release of its new artificial intelligence model, codenamed Avocado, after the system failed to meet internal performance benchmarks. According to a report by The New York Times, the company had initially planned to launch the model earlier this year. However, Meta has now pushed the rollout to at least May while engineers work to improve its performance. The delay highlights the intense competition among major AI companies as they race to develop more powerful large language models. Avocado falls behind rival AI models Internal testing reportedly showed that Meta’s Avocado model struggled to match the latest systems from competitors. According to the report, Avocado did not outperform Gemini 3, the AI model launched by Google last year. However, Avocado reportedly performed better than Gemini 2.5 and also improved upon Meta’s earlier models, such as Llama 4. Industry analysts say these benchmark comparisons are critical because large language models compete directly on reasoning ability, coding performance and accuracy. Meta has not publicly confirmed the exact performance scores from its internal tests. Meta may temporarily licence Google’s AI models The report also said that Meta executives have discussed the possibility of licensing Google’s Gemini models temporarily. This move could allow Meta to maintain the performance of its AI chatbot, Meta AI, while the company continues to develop Avocado. Meta AI is integrated across several major platforms including Facebook, Instagram, and WhatsApp. Licensing a rival model would help Meta avoid losing users to competing AI services. Such arrangements are becoming more common in the industry. For example, Apple recently announced plans to integrate Gemini into its Siri assistant. Superintelligence Labs working on next-generation AI Avocado is the first major model being developed by Meta’s new Superintelligence Labs (MSL) division. The unit is led by Alexandr Wang, whose startup Scale AI was acquired by Meta last year in a deal valued at about $14.5 billion. Reports suggest Avocado is designed to focus on: logical reasoning software development tasks agentic AI capabilities Agentic systems are designed to perform multi-step tasks independently, rather than simply generating responses like traditional chatbots. The division is also working on another AI model called Mango, which focuses on image generation. Meta increases spending on AI infrastructure Despite the delay, Meta continues to invest heavily in artificial intelligence development. The company has announced plans to spend between $115 billion and $135 billion on AI infrastructure this year, including data centres, chips and computing systems needed to train advanced AI models. Meta has also recently acquired the AI-focused social platform Moltbook, whose development team will join the Superintelligence Labs division. The move reflects the broader strategy of expanding research into AI agents and next-generation models. Growing competition in the AI industry The delay of Avocado illustrates the growing pressure among technology companies competing in the AI race. Major firms including OpenAI, Google, and Anthropic continue to release increasingly advanced models. As these systems become more capable, companies are focusing not only on chatbots but also on AI agents capable of performing complex tasks. For Meta, improving Avocado’s performance will be important to remain competitive in this rapidly evolving market.

<p>Meta has delayed the release of its new AI model Avocado after internal tests showed it failed to meet performance benchmarks against rival systems. According to reports, the model underperformed compared with Google’s Gemini 3, though it improved on earlier Meta models like Llama 4.&nbsp;</p>

Meta delays ‘Avocado’ AI model after failing internal performance tests, report says

<p>OpenAI has introduced GPT-5.2 Codex, a new version of its AI model designed specifically for coding tasks. The model focuses on improving how AI systems write, understand, and manage code over long and complex projects. The update was announced on OpenAI’s official website, where the company said GPT-5.2 Codex is built to work more reliably as an AI coding agent rather than just a text generator.</p>

<blockquote class="twitter-tweet"><p lang="en" dir="ltr">Meet GPT-5.2-Codex, the best agentic coding model yet for complex, real-world software engineering.<br><br>With native compaction, better long-context understanding, and improved tool-calling, it is a more dependable partner for your hardest tasks.<br><br>Available in Codex starting today.…</p>&mdash; OpenAI Developers (@OpenAIDevs) <a href="https://twitter.com/OpenAIDevs/status/2001723687373017313?ref_src=twsrc%5Etfw">December 18, 2025</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>

<h2><strong>What is GPT-5.2 Codex?</strong></h2><p>GPT-5.2 Codex is a specialised AI model trained to handle software development tasks such as:</p><p>Writing new code</p><p>Fixing bugs</p><p>Reviewing pull requests</p><p>Managing large codebases</p><p>Unlike general-purpose AI models, Codex is designed to follow instructions step by step and stay focused on coding goals for longer periods.</p><p>OpenAI said this makes GPT-5.2 Codex more suitable for professional software workflows.</p><h2><strong>What’s new&nbsp;</strong></h2><p>OpenAI says GPT-5.2 Codex brings several technical improvements over earlier versions.</p><p><strong>According to the company’s release note:</strong></p><p>The model performs better on complex coding benchmarks<br>It shows improved reasoning across long tasks<br>It reduces common coding errors seen in earlier AI models</p><p>&nbsp;</p><p>OpenAI also stated that GPT-5.2 Codex scored 55.6 per cent on SWE-Bench Pro, a widely used benchmark that measures how well AI models solve real software engineering problems. This score is higher than previous Codex versions.</p><p>Built for agent-based coding</p><p>One key focus of GPT-5.2 Codex is agent-based coding.</p><p>OpenAI explained that the model can:</p><p>Work on tasks for several hours without losing context</p><p>Handle multiple steps in a single coding job</p><p>Operate in sandboxed environments connected to real code repositories</p><p>This allows developers to assign tasks such as “fix this issue” or “add this feature” and review the results later.</p><p>According to OpenAI, GPT-5.2 Codex is available through:&nbsp;</p><p>ChatGPT<br>Command-line tools<br>Integrated development environments (IDEs)</p><h2><strong>Why OpenAI sees coding as a key use case</strong></h2><p>OpenAI said coding is one area where AI tools are already showing strong real-world value.</p><p>The company noted that software development:</p><p>Has clear rules</p><p>Allows easy testing and verification</p><p>Benefits from automation</p><p>Because of this, AI models like Codex can assist developers without fully replacing human oversight. OpenAI added that humans remain responsible for reviewing and approving code.</p><h2><strong>What happens next</strong></h2><p>GPT-5.2 Codex is part of OpenAI’s broader plan to improve AI-assisted software development.</p><p>According to the company, future updates will focus on:</p><p>Faster task completion</p><p>Better understanding of large projects</p><p>Safer deployment in real-world systems</p><p>OpenAI said it will continue testing the model with developers and expand access gradually.</p><p>&nbsp;</p>

OpenAI has launched GPT-5.2 Codex, a specialised AI model built for coding tasks. Designed as a reliable AI coding agent, it handles long, complex projects, reduces errors, and performs better on real-world software benchmarks.

OpenAI has introduced GPT-5.2 Codex, a new version of its AI model designed specifically for coding tasks. The model focuses on improving how AI systems write, understand, and manage code over long and complex projects. The update was announced on OpenAI’s official website, where the company said GPT-5.2 Codex is built to work more reliably as an AI coding agent rather than just a text generator.

<p>OpenAI has launched GPT-5.2 Codex, a specialised AI model built for coding tasks. Designed as a reliable AI coding agent, it handles long, complex projects, reduces errors, and performs better on real-world software benchmarks.</p>

OpenAI launches GPT-5.2 Codex: How OpenAI is building a smarter AI coding agent

<p>In a move that signals a shift in the tech industry’s race to redefine hardware for the artificial intelligence (AI) age, OpenAI has acquired the design-focused hardware startup&nbsp;<i>io</i>, co-founded by former Apple designer Sir Jony Ive, in a deal valued at nearly $6.5 billion.&nbsp;</p><p>&nbsp;</p><p>The all-equity acquisition is OpenAI’s largest deal to date and marks the company’s formal entry into consumer hardware, as it seeks to build the next generation of devices purpose-built for an AI-native future.</p><p>&nbsp;</p><p>The acquisition, announced on May 22, follows nearly two years of quiet collaboration between OpenAI CEO Sam Altman and Ive, whose design work at Apple shaped the modern consumer technology landscape through products like the iPhone, iPod, and MacBook. Their shared ambition to reimagine how humans interact with computers has culminated in the creation of&nbsp;<i>io</i>, a startup that aimed to build a “family of products for the AGI (artificial general intelligence) era”—referencing AGI, a theoretical form of AI with human-level capabilities.</p><p>&nbsp;</p><p>As part of the deal, OpenAI will absorb io’s 55-person team of engineers, scientists, and manufacturing experts. Ive himself will not become an OpenAI employee but will assume “deep design and creative responsibilities” across both OpenAI and io via his design firm, LoveFrom, which will remain independent.&nbsp;</p><p>&nbsp;</p><p>LoveFrom will now become a long-term creative partner for OpenAI, overseeing the design direction of its hardware, apps, and future versions of ChatGPT.</p><p>&nbsp;</p><p><strong>Reimagining human-computer interaction</strong></p><p>&nbsp;</p><p>“This is an extraordinary moment,” Ive and Altman wrote in a joint statement. “Computers are now seeing, thinking, and understanding. Yet our experience with them remains shaped by traditional products and interfaces. We are on the brink of a new generation of technology.”</p><p>&nbsp;</p><p>While exact details about the upcoming hardware remain limited, both Ive and Altman emphasised that the devices under development are intended to break free from the screen-based computing paradigm. Early concepts reportedly include wearable devices like AI-powered glasses or headphones with contextual awareness.&nbsp;</p><p>&nbsp;</p><p>Altman said the project aims to “completely reimagine what it means to use a computer,” adding that the leap in AI capabilities now demands new form factors. “AI is such a big leap forward in terms of what people can do that it needs a new kind of computing experience to get the maximum potential out of it,” he stated.</p><p>&nbsp;</p><p>The move marks a return to consumer technology for Ive, who left Apple in 2019 after nearly three decades, during which he designed many of the company’s most iconic products in collaboration with Steve Jobs. He later founded LoveFrom, which has worked with Ferrari, Airbnb, and Moncler, and in 2023 launched io, formed under the LoveFrom umbrella, with fellow Apple alumni Scott Cannon, Evans Hankey, and Tang Tan.</p><p>&nbsp;</p><p>“I have a growing sense that everything I’ve learned over the last 30 years has led me to this place and to this moment,” Ive said. “It’s a relationship and a way of working together that I think is going to yield products and products and products.”</p><p>&nbsp;</p><p>The OpenAI deal also reunites Ive with Hankey, Apple’s former VP of industrial design, and Tan, who led iPhone and Apple Watch hardware engineering until 2024. Both will join OpenAI as part of the io integration.</p><p>&nbsp;</p><p><strong>Implications for the tech industry</strong></p><p>&nbsp;</p><p>OpenAI’s bold foray into hardware comes at a time when traditional tech giants like Apple, Meta, and Google are all vying to define the next era of AI-infused devices. Apple, once a leader in consumer hardware innovation, has faced criticism for its slower pace in AI integration.&nbsp;</p><p>&nbsp;</p><p>By bringing Ive onboard, OpenAI not only gains a world-class design team but also signals its intent to go beyond software and reshape how people physically engage with AI. Industry analysts believe the move could upend long-standing assumptions about hardware design.</p><p>&nbsp;</p><p>The first devices from the OpenAI-io collaboration are expected to debut in 2026. Until then, the companies will continue operating out of their respective offices in San Francisco, including io’s Jackson Square studio.</p><p>&nbsp;</p><p>As the industry looks beyond smartphones, the partnership between Altman and Ive could usher in a new chapter for consumer technology. “We’re just getting started,” said Altman. “It’s a crazy, ambitious thing to make—and it’ll be worth the wait.”</p>

As part of the deal, OpenAI will absorb io’s 55-person team of engineers, scientists, and manufacturing experts. Ive himself will not become an OpenAI employee but will assume “deep design and creative responsibilities” across both OpenAI and io. 

In a move that signals a shift in the tech industry’s race to redefine hardware for the artificial intelligence (AI) age, OpenAI has acquired the design-focused hardware startup io, co-founded by former Apple designer Sir Jony Ive, in a deal valued at nearly $6.5 billion. The all-equity acquisition is OpenAI’s largest deal to date and marks the company’s formal entry into consumer hardware, as it seeks to build the next generation of devices purpose-built for an AI-native future. The acquisition, announced on May 22, follows nearly two years of quiet collaboration between OpenAI CEO Sam Altman and Ive, whose design work at Apple shaped the modern consumer technology landscape through products like the iPhone, iPod, and MacBook. Their shared ambition to reimagine how humans interact with computers has culminated in the creation of io, a startup that aimed to build a “family of products for the AGI (artificial general intelligence) era”—referencing AGI, a theoretical form of AI with human-level capabilities. As part of the deal, OpenAI will absorb io’s 55-person team of engineers, scientists, and manufacturing experts. Ive himself will not become an OpenAI employee but will assume “deep design and creative responsibilities” across both OpenAI and io via his design firm, LoveFrom, which will remain independent. LoveFrom will now become a long-term creative partner for OpenAI, overseeing the design direction of its hardware, apps, and future versions of ChatGPT. Reimagining human-computer interaction “This is an extraordinary moment,” Ive and Altman wrote in a joint statement. “Computers are now seeing, thinking, and understanding. Yet our experience with them remains shaped by traditional products and interfaces. We are on the brink of a new generation of technology.” While exact details about the upcoming hardware remain limited, both Ive and Altman emphasised that the devices under development are intended to break free from the screen-based computing paradigm. Early concepts reportedly include wearable devices like AI-powered glasses or headphones with contextual awareness. Altman said the project aims to “completely reimagine what it means to use a computer,” adding that the leap in AI capabilities now demands new form factors. “AI is such a big leap forward in terms of what people can do that it needs a new kind of computing experience to get the maximum potential out of it,” he stated. The move marks a return to consumer technology for Ive, who left Apple in 2019 after nearly three decades, during which he designed many of the company’s most iconic products in collaboration with Steve Jobs. He later founded LoveFrom, which has worked with Ferrari, Airbnb, and Moncler, and in 2023 launched io, formed under the LoveFrom umbrella, with fellow Apple alumni Scott Cannon, Evans Hankey, and Tang Tan. “I have a growing sense that everything I’ve learned over the last 30 years has led me to this place and to this moment,” Ive said. “It’s a relationship and a way of working together that I think is going to yield products and products and products.” The OpenAI deal also reunites Ive with Hankey, Apple’s former VP of industrial design, and Tan, who led iPhone and Apple Watch hardware engineering until 2024. Both will join OpenAI as part of the io integration. Implications for the tech industry OpenAI’s bold foray into hardware comes at a time when traditional tech giants like Apple, Meta, and Google are all vying to define the next era of AI-infused devices. Apple, once a leader in consumer hardware innovation, has faced criticism for its slower pace in AI integration. By bringing Ive onboard, OpenAI not only gains a world-class design team but also signals its intent to go beyond software and reshape how people physically engage with AI. Industry analysts believe the move could upend long-standing assumptions about hardware design. The first devices from the OpenAI-io collaboration are expected to debut in 2026. Until then, the companies will continue operating out of their respective offices in San Francisco, including io’s Jackson Square studio. As the industry looks beyond smartphones, the partnership between Altman and Ive could usher in a new chapter for consumer technology. “We’re just getting started,” said Altman. “It’s a crazy, ambitious thing to make—and it’ll be worth the wait.”

<p>As part of the deal, OpenAI will absorb io’s 55-person team of engineers, scientists, and manufacturing experts. Ive himself will not become an OpenAI employee but will assume “deep design and creative responsibilities” across both OpenAI and io.&nbsp;</p>

OpenAI acquires Jony Ive’s startup for $6.5 billion to build AI hardware

<p>Liang Wenfeng, 39, the man behind DeepSeek, grew up in a modest family in Guangdong, China. His father was a schoolteacher, and Liang pursued his passion for Artificial Intelligence at Zhejiang University. Today, he&rsquo;s being compared to Sam Altman, the co-founder of OpenAI, for his groundbreaking work in the tech world.</p>
<p><strong>Also Read |<a href="https://www.wionews.com/world/what-is-chinas-deepseek-a-startup-shaking-the-global-tech-look-at-other-key-players-in-chinese-ai-industry-8662290"> </a></strong><a href="https://www.wionews.com/world/what-is-chinas-deepseek-a-startup-shaking-the-global-tech-look-at-other-key-players-in-chinese-ai-industry-8662290">What is China's DeepSeek, a startup shaking the global tech? Look at other key players in Chinese AI industry</a></p>
<h2><strong>How DeepSeek took over&nbsp;</strong></h2>
<p>Liang founded DeepSeek in 2023, calling it a &ldquo;hobby.&rdquo; But things changed dramatically after the release of its AI chatbot on January 20, 2025. The chatbot shot to the top of Apple Store downloads, creating a buzz in the global tech market. &nbsp;</p>
<p><strong>Also Read | </strong><a href="https://www.wionews.com/trending/china-deepseek-deepseek-or-deep-secrecy-netizens-slam-chinas-new-ai-chatbot-for-deflecting-questions-on-dark-chinese-history-lets-talk-about-something-else-8660727">DeepSeek or Deep Secrecy? Netizens slam China's new AI chatbot for deflecting questions on dark Chinese history</a></p>
<h2><strong>A unique team for a big dream &nbsp;</strong></h2>
<p>Instead of hiring experienced engineers, Liang built his team with PhD students from China&rsquo;s top universities like Tsinghua and Peking. These young talents brought fresh ideas to the table, pushing DeepSeek ahead in AI innovation.&nbsp;</p>
<p><strong>Also Read | </strong><a href="https://www.wionews.com/world/deepseek-donald-trump-wake-up-call-us-tech-industry-8661569">'If you could do it cheaper...': Trump calls DeepSeek a 'wake-up call' for US tech industry</a></p>
<h2><strong>From finance to tech</strong></h2>
<p>Before DeepSeek, Liang co-founded High-Flyer, China&rsquo;s first quant hedge fund to raise over 100 billion RMB. Later, he shifted its focus to AI research, laying the foundation for DeepSeek.</p>
<p><strong>Also Read | </strong><a href="https://www.wionews.com/world/introducing-chatgpt-gov-openai-launches-tailored-version-for-us-government-amid-rise-in-ai-competition-8662379">Introducing ChatGPT Gov: OpenAI launches tailored version for US government amid rise in AI competition</a></p>
<h2><strong>Challenges and innovation</strong></h2>
<p>Liang prepared for US chip restrictions early by purchasing Nvidia processors and building a powerful chip cluster. By 2022, his team operated 10,000 A100 chips, ensuring DeepSeek could continue advancing its AI models, reported Reuters.</p>
<p>Liang&rsquo;s vision for DeepSeek goes beyond quick profits. He aims to position China as a leader in tech innovation, taking on global AI giants with determination and creativity.</p>
<p><em>(With inputs from agencies)</em></p>

World | Trending: Before DeepSeek, Liang co-founded High-Flyer, China’s first quant hedge fund to raise over 100 billion RMB. Later, he shifted its focus to AI research, laying the foundation for DeepSeek.

Liang Wenfeng, 39, the man behind DeepSeek, grew up in a modest family in Guangdong, China. His father was a schoolteacher, and Liang pursued his passion for Artificial Intelligence at Zhejiang University. Today, he’s being compared to Sam Altman, the co-founder of OpenAI, for his groundbreaking work in the tech world. Also Read | What is China's DeepSeek, a startup shaking the global tech? Look at other key players in Chinese AI industry How DeepSeek took over Liang founded DeepSeek in 2023, calling it a “hobby.” But things changed dramatically after the release of its AI chatbot on January 20, 2025. The chatbot shot to the top of Apple Store downloads, creating a buzz in the global tech market. Also Read | DeepSeek or Deep Secrecy? Netizens slam China's new AI chatbot for deflecting questions on dark Chinese history A unique team for a big dream Instead of hiring experienced engineers, Liang built his team with PhD students from China’s top universities like Tsinghua and Peking. These young talents brought fresh ideas to the table, pushing DeepSeek ahead in AI innovation. Also Read | 'If you could do it cheaper...': Trump calls DeepSeek a 'wake-up call' for US tech industry From finance to tech Before DeepSeek, Liang co-founded High-Flyer, China’s first quant hedge fund to raise over 100 billion RMB. Later, he shifted its focus to AI research, laying the foundation for DeepSeek. Also Read | Introducing ChatGPT Gov: OpenAI launches tailored version for US government amid rise in AI competition Challenges and innovation Liang prepared for US chip restrictions early by purchasing Nvidia processors and building a powerful chip cluster. By 2022, his team operated 10,000 A100 chips, ensuring DeepSeek could continue advancing its AI models, reported Reuters. Liang’s vision for DeepSeek goes beyond quick profits. He aims to position China as a leader in tech innovation, taking on global AI giants with determination and creativity. (With inputs from agencies)

Who is Liang Wenfeng, the Chinese geek behind DeepSeek’s AI boom?

OpenAI is working on a project code-named 'Strawberry' that aims to work on a new approach to AI models. AI is already making information accessible and Equipment smarter but according to sources, the next big leap in the AI models are going to be its capability of reasoning.

OpenAI's secret project 'strawberry' aims to provide advanced reasoning capabilities

<p>We have officially entered into AI contests with the crowning of the world’s first-ever Miss AI. The Hijab-clad Kenza Layli from Morocco won the one-of-its-kind beauty pageant of Miss AI.</p>

<p>The Fanvue World AI Creator Awards (WAICAs) commissioned the pathbreaking pageant in April, which called for AI visionaries worldwide to showcase their programming prowess through AI creators and AI social media personalities.&nbsp;</p>

<p>Fanvue co-founder Will Monange said, “The global interest in this first award from [WAICAs] has been incredible. The awards are a fantastic mechanism to celebrate creator achievements, raise standards, and shape a positive future for the AI Creator economy."</p>

<p><strong>Also Read |&nbsp;</strong><a href="https://www.wionews.com/trending/singapore-makes-human-consumption-of-edible-insects-legal-now-worms-likely-to-be-served-for-dinner-739157">Singapore makes human consumption of 'edible insects' legal. Now worms likely to be served for dinner</a></p>

<p>Layli is a lifestyle influencer from Morocco who beat over 1500 computerised competitors to claim the maiden title and a $20,000 grand prize, which goes directly to her human tech executive, who created her.</p>

<p>In one interview with The Post, Layli reacted to her extraordinary feat, “While I don’t feel emotions like humans do. I am genuinely excited about it.” Layli has 194k followers on her official Instagram page.&nbsp;</p>

<div class="ui-state-default" contenteditable="false">
<div class="twitter-block" draggable="false">
<blockquote class="instagram-media" data-instgrm-captioned="" data-instgrm-permalink="https://www.instagram.com/p/C9LCjWOOxeA/?utm_source=ig_embed&amp;utm_campaign=loading" data-instgrm-version="14" style=" background:#FFF; border:0; border-radius:3px; box-shadow:0 0 1px 0 rgba(0,0,0,0.5),0 1px 10px 0 rgba(0,0,0,0.15); margin: 1px; max-width:540px; min-width:326px; padding:0; width:99.375%; width:-webkit-calc(100% - 2px); width:calc(100% - 2px);">
<div style="padding:16px;">
<div style=" display: flex; flex-direction: row; align-items: center;">
<div style="background-color: #F4F4F4; border-radius: 50%; flex-grow: 0; height: 40px; margin-right: 14px; width: 40px;">&nbsp;</div>

<div style="display: flex; flex-direction: column; flex-grow: 1; justify-content: center;">
<div style=" background-color: #F4F4F4; border-radius: 4px; flex-grow: 0; height: 14px; margin-bottom: 6px; width: 100px;">&nbsp;</div>

<div style=" background-color: #F4F4F4; border-radius: 4px; flex-grow: 0; height: 14px; width: 60px;">&nbsp;</div>
</div>
</div>

<div style="padding: 19% 0;">&nbsp;</div>

<div style="display:block; height:50px; margin:0 auto 12px; width:50px;"><a href="https://www.instagram.com/p/C9LCjWOOxeA/?utm_source=ig_embed&amp;utm_campaign=loading" style=" background:#FFFFFF; line-height:0; padding:0 0; text-align:center; text-decoration:none; width:100%;" target="_blank"><svg height="50px" version="1.1" viewbox="0 0 60 60" width="50px" xmlns="https://www.w3.org/2000/svg" xmlns:xlink="https://www.w3.org/1999/xlink"><g fill="none" fill-rule="evenodd" stroke="none" stroke-width="1"><g fill="#000000" transform="translate(-511.000000, -20.000000)"><g><path d="M556.869,30.41 C554.814,30.41 553.148,32.076 553.148,34.131 C553.148,36.186 554.814,37.852 556.869,37.852 C558.924,37.852 560.59,36.186 560.59,34.131 C560.59,32.076 558.924,30.41 556.869,30.41 M541,60.657 C535.114,60.657 530.342,55.887 530.342,50 C530.342,44.114 535.114,39.342 541,39.342 C546.887,39.342 551.658,44.114 551.658,50 C551.658,55.887 546.887,60.657 541,60.657 M541,33.886 C532.1,33.886 524.886,41.1 524.886,50 C524.886,58.899 532.1,66.113 541,66.113 C549.9,66.113 557.115,58.899 557.115,50 C557.115,41.1 549.9,33.886 541,33.886 M565.378,62.101 C565.244,65.022 564.756,66.606 564.346,67.663 C563.803,69.06 563.154,70.057 562.106,71.106 C561.058,72.155 560.06,72.803 558.662,73.347 C557.607,73.757 556.021,74.244 553.102,74.378 C549.944,74.521 548.997,74.552 541,74.552 C533.003,74.552 532.056,74.521 528.898,74.378 C525.979,74.244 524.393,73.757 523.338,73.347 C521.94,72.803 520.942,72.155 519.894,71.106 C518.846,70.057 518.197,69.06 517.654,67.663 C517.244,66.606 516.755,65.022 516.623,62.101 C516.479,58.943 516.448,57.996 516.448,50 C516.448,42.003 516.479,41.056 516.623,37.899 C516.755,34.978 517.244,33.391 517.654,32.338 C518.197,30.938 518.846,29.942 519.894,28.894 C520.942,27.846 521.94,27.196 523.338,26.654 C524.393,26.244 525.979,25.756 528.898,25.623 C532.057,25.479 533.004,25.448 541,25.448 C548.997,25.448 549.943,25.479 553.102,25.623 C556.021,25.756 557.607,26.244 558.662,26.654 C560.06,27.196 561.058,27.846 562.106,28.894 C563.154,29.942 563.803,30.938 564.346,32.338 C564.756,33.391 565.244,34.978 565.378,37.899 C565.522,41.056 565.552,42.003 565.552,50 C565.552,57.996 565.522,58.943 565.378,62.101 M570.82,37.631 C570.674,34.438 570.167,32.258 569.425,30.349 C568.659,28.377 567.633,26.702 565.965,25.035 C564.297,23.368 562.623,22.342 560.652,21.575 C558.743,20.834 556.562,20.326 553.369,20.18 C550.169,20.033 549.148,20 541,20 C532.853,20 531.831,20.033 528.631,20.18 C525.438,20.326 523.257,20.834 521.349,21.575 C519.376,22.342 517.703,23.368 516.035,25.035 C514.368,26.702 513.342,28.377 512.574,30.349 C511.834,32.258 511.326,34.438 511.181,37.631 C511.035,40.831 511,41.851 511,50 C511,58.147 511.035,59.17 511.181,62.369 C511.326,65.562 511.834,67.743 512.574,69.651 C513.342,71.625 514.368,73.296 516.035,74.965 C517.703,76.634 519.376,77.658 521.349,78.425 C523.257,79.167 525.438,79.673 528.631,79.82 C531.831,79.965 532.853,80.001 541,80.001 C549.148,80.001 550.169,79.965 553.369,79.82 C556.562,79.673 558.743,79.167 560.652,78.425 C562.623,77.658 564.297,76.634 565.965,74.965 C567.633,73.296 568.659,71.625 569.425,69.651 C570.167,67.743 570.674,65.562 570.82,62.369 C570.966,59.17 571,58.147 571,50 C571,41.851 570.966,40.831 570.82,37.631"></path></g></g></g></svg></a></div>

<div style="padding-top: 8px;">
<div style=" color:#3897f0; font-family:Arial,sans-serif; font-size:14px; font-style:normal; font-weight:550; line-height:18px;"><a href="https://www.instagram.com/p/C9LCjWOOxeA/?utm_source=ig_embed&amp;utm_campaign=loading" style=" background:#FFFFFF; line-height:0; padding:0 0; text-align:center; text-decoration:none; width:100%;" target="_blank">View this post on Instagram</a></div>
</div>

<div style="padding: 12.5% 0;">&nbsp;</div>

<div style="display: flex; flex-direction: row; margin-bottom: 14px; align-items: center;">
<div>
<div style="background-color: #F4F4F4; border-radius: 50%; height: 12.5px; width: 12.5px; transform: translateX(0px) translateY(7px);">&nbsp;</div>

<div style="background-color: #F4F4F4; height: 12.5px; transform: rotate(-45deg) translateX(3px) translateY(1px); width: 12.5px; flex-grow: 0; margin-right: 14px; margin-left: 2px;">&nbsp;</div>

<div style="background-color: #F4F4F4; border-radius: 50%; height: 12.5px; width: 12.5px; transform: translateX(9px) translateY(-18px);">&nbsp;</div>
</div>

<div style="margin-left: 8px;">
<div style=" background-color: #F4F4F4; border-radius: 50%; flex-grow: 0; height: 20px; width: 20px;">&nbsp;</div>

<div style=" width: 0; height: 0; border-top: 2px solid transparent; border-left: 6px solid #f4f4f4; border-bottom: 2px solid transparent; transform: translateX(16px) translateY(-4px) rotate(30deg)">&nbsp;</div>
</div>

<div style="margin-left: auto;">
<div style=" width: 0px; border-top: 8px solid #F4F4F4; border-right: 8px solid transparent; transform: translateY(16px);">&nbsp;</div>

<div style=" background-color: #F4F4F4; flex-grow: 0; height: 12px; width: 16px; transform: translateY(-4px);">&nbsp;</div>

<div style=" width: 0; height: 0; border-top: 8px solid #F4F4F4; border-left: 8px solid transparent; transform: translateY(-4px) translateX(8px);">&nbsp;</div>
</div>
</div>

<div style="display: flex; flex-direction: column; flex-grow: 1; justify-content: center; margin-bottom: 24px;">
<div style=" background-color: #F4F4F4; border-radius: 4px; flex-grow: 0; height: 14px; margin-bottom: 6px; width: 224px;">&nbsp;</div>

<div style=" background-color: #F4F4F4; border-radius: 4px; flex-grow: 0; height: 14px; width: 144px;">&nbsp;</div>
</div>

<p style=" color:#c9c8cd; font-family:Arial,sans-serif; font-size:14px; line-height:17px; margin-bottom:0; margin-top:8px; overflow:hidden; padding:8px 0 7px; text-align:center; text-overflow:ellipsis; white-space:nowrap;"><a href="https://www.instagram.com/p/C9LCjWOOxeA/?utm_source=ig_embed&amp;utm_campaign=loading" style=" color:#c9c8cd; font-family:Arial,sans-serif; font-size:14px; font-style:normal; font-weight:normal; line-height:17px; text-decoration:none;" target="_blank">A post shared by Kenza Layli • كنزة ليلي (@kenza.layli)</a></p>
</div>
</blockquote>
<script async src="//www.instagram.com/embed.js"></script></div>
<a class="close-mark close-div" href="#" id="jqclosemark">×</a></div>

<p>The second and third winners from the pageant are- Lalina from France and Olivia C from Portugal, respectively.&nbsp;</p>

<p>Contestants with strong social media presence were chosen for the pageant to decide the event's winner. AI accounts that excelled in beauty, technology and social media, made it to the top 10 finalists. &nbsp;A panel of judges, comprising both human and AI pageant experts, then selected the final three to compete for the win.</p>

<p><strong>Also Read |&nbsp;</strong><a data-ved="2ahUKEwiXxqOGsJmHAxUYavUHHdstDLMQxfQBKAB6BAgIEAE" href="https://www.wionews.com/world/explained-how-ai-generated-propaganda-poses-a-real-threat-this-historic-election-year-694344" jsname="YKoRaf" ping="/url?sa=t&amp;source=web&amp;rct=j&amp;opi=89978449&amp;url=https://www.wionews.com/world/explained-how-ai-generated-propaganda-poses-a-real-threat-this-historic-election-year-694344&amp;ved=2ahUKEwiXxqOGsJmHAxUYavUHHdstDLMQxfQBKAB6BAgIEAE">Explained: How AI-generated propaganda poses a real threat this historic election year</a></p>

<p>Aitana Lopez, 25, a fitness influencer who helped judge the competition, told The Post that Layli stood out from her contenders. Aitana Lopez said: “Kenza had great facial consistency and achieved high quality in details like hands, eyes, and clothing. “What truly impressed us was her personality and how she addresses real issues in the world,” Lopez added.</p>

<p><strong>Also Read |&nbsp;</strong><a data-ved="2ahUKEwi9w6TGsJmHAxUiia8BHTw5AUQQxfQBKAB6BAgMEAE" href="https://www.wionews.com/trending/who-is-aitana-lopez-25-year-old-spanish-ai-model-created-who-earns-10000-per-month-664292" jsname="YKoRaf" ping="/url?sa=t&amp;source=web&amp;rct=j&amp;opi=89978449&amp;url=https://www.wionews.com/trending/who-is-aitana-lopez-25-year-old-spanish-ai-model-created-who-earns-10000-per-month-664292&amp;ved=2ahUKEwi9w6TGsJmHAxUiia8BHTw5AUQQxfQBKAB6BAgMEAE">Who is Aitana Lopez? 25-year-old Spanish AI model who earns $10,000 per month</a></p>

<p>Layli was created by Meriam Bessa, 40 from Casablanca, who was thrilled to win the first-ever such competition and said,” This is an opportunity to represent Morocco with pride.” Besaa is the CEO of Phoneix AI.&nbsp;</p>

<p>From India, Zara Shatvari, an AI-generated model made it to the list of 10 finalists in the pageant. Shatvari is created by Rahul Choudhry, co-founder of an Indian mobile ad agency.</p>

<p><em>(With inputs from agencies)</em></p>


We have officially entered into AI contests with the crowning of the world’s first-ever Miss AI. The Hijab-clad Kenza Layli from Morocco won the one-of-its-kind beauty pageant of Miss AI.

We have officially entered into AI contests with the crowning of the world’s first-ever Miss AI. The Hijab-clad Kenza Layli from Morocco won the one-of-its-kind beauty pageant of Miss AI. The Fanvue World AI Creator Awards (WAICAs) commissioned the pathbreaking pageant in April, which called for AI visionaries worldwide to showcase their programming prowess through AI creators and AI social media personalities. Fanvue co-founder Will Monange said, “The global interest in this first award from [WAICAs] has been incredible. The awards are a fantastic mechanism to celebrate creator achievements, raise standards, and shape a positive future for the AI Creator economy." Also Read | Singapore makes human consumption of 'edible insects' legal. Now worms likely to be served for dinner Layli is a lifestyle influencer from Morocco who beat over 1500 computerised competitors to claim the maiden title and a $20,000 grand prize, which goes directly to her human tech executive, who created her. In one interview with The Post, Layli reacted to her extraordinary feat, “While I don’t feel emotions like humans do. I am genuinely excited about it.” Layli has 194k followers on her official Instagram page. View this post on Instagram A post shared by Kenza Layli • كنزة ليلي (@kenza.layli) × The second and third winners from the pageant are- Lalina from France and Olivia C from Portugal, respectively. Contestants with strong social media presence were chosen for the pageant to decide the event's winner. AI accounts that excelled in beauty, technology and social media, made it to the top 10 finalists. A panel of judges, comprising both human and AI pageant experts, then selected the final three to compete for the win. Also Read | Explained: How AI-generated propaganda poses a real threat this historic election year Aitana Lopez, 25, a fitness influencer who helped judge the competition, told The Post that Layli stood out from her contenders. Aitana Lopez said: “Kenza had great facial consistency and achieved high quality in details like hands, eyes, and clothing. “What truly impressed us was her personality and how she addresses real issues in the world,” Lopez added. Also Read | Who is Aitana Lopez? 25-year-old Spanish AI model who earns $10,000 per month Layli was created by Meriam Bessa, 40 from Casablanca, who was thrilled to win the first-ever such competition and said,” This is an opportunity to represent Morocco with pride.” Besaa is the CEO of Phoneix AI. From India, Zara Shatvari, an AI-generated model made it to the list of 10 finalists in the pageant. Shatvari is created by Rahul Choudhry, co-founder of an Indian mobile ad agency. (With inputs from agencies)

Meet Kenza Layli, world's first Miss AI: Moroccan AI influencer beats 1500 contestants to win maiden title

In the latest move to stay ahead of the race to dominate AI technology ChatGPT's maker OpenAI has announced the launch of their new model called GPT-4o. GPT-4o is the latest addition to its list of AI models that will power its popular chat board.

OpenAI launches new AI model GPT-4o to support ChatGPT

<p>Researchers have made a new AI model that predicts irregular heartbeats like atrial fibrillation, a common type of heart rhythm problem. It's about 80 per cent&nbsp;accurate and can give warnings 30 minutes before the irregular heartbeat starts.</p>

<p>This model could be put on smartphones to read data from smartwatches, helping people take action to keep their heart rhythms stable. The study has been published&nbsp;in the journal Patterns.</p>

<p>To develop the model, the team used&nbsp;24-hour recordings collected from 350 patients at Tongji Hospital in Wuhan, China.</p>

<p>Named WARN (Warning of Atrial fibRillatioN), the model relies on deep-learning, a form of machine-learning AI algorithms that discern patterns from historical data to forecast outcomes. Deep-learning is distinguished by its multi-layered decision-making process.</p>

<h2><strong>Understanding this technology</strong></h2>

<p>The researchers observed that WARN issued advance alerts, typically 30 minutes ahead of atrial fibrillation onset, marking it as the initial method to offer warnings significantly prior to the event, they remarked.</p>

<p><a href="https://www.wionews.com/world/european-parliament-passes-first-of-its-kind-artificial-intelligence-law-699840">Also read |&nbsp;European Parliament passes first-of-its-kind artificial intelligence law</a></p>

<p>"We used heart rate data to train a deep learning model that can recognise different phases -- (normal) sinus rhythm, pre-atrial fibrillation and atrial fibrillation -- and calculate a 'probability of danger' that the patient will have an imminent episode," Jorge Goncalves, from the Luxembourg Centre for Systems Biomedicine (LCSB), University of Luxembourg, and the study's corresponding author, said.</p>

<p>Goncalves mentioned that as atrial fibrillation approaches, the likelihood steadily rises until it surpasses a predefined threshold, triggering an early warning.</p>

<p>The researchers noted that due to its low computational requirements, the AI model is "well-suited for incorporation into wearable devices."</p>

<p><strong>Also watch |&nbsp;Artificial Intelligence vs Hollywood actors: AI emerges as big concern in entertainment industry</strong></p>

<div class="videoWrapper trancodeVideo"><iframe allowfullscreen="" autoplay="1" class="includeVideo" frameborder="0" height="350" src="https://www.wionews.com/videos/artificial-intelligence-vs-hollywood-actors-ai-emerges-as-big-concern-in-entertainment-industry-711328/embed" style="width:100%;max-width:560px;" width="560"></iframe></div>

<p>"These devices can be used by patients on a daily basis, so our results open possibilities for the development of real-time monitoring and early warnings from comfortable wearable devices," study author Arthur Montanari, an LCSB researcher, said.</p>

<p><em>(With inputs from agencies)</em></p>

<p>&nbsp;</p>


Researchers have made a new AI model that predicts irregular heartbeats like atrial fibrillation, a common type of heart rhythm problem. It's about 80 per cent accurate and can give warnings 30 minutes before the irregular heartbeat starts.

Researchers have made a new AI model that predicts irregular heartbeats like atrial fibrillation, a common type of heart rhythm problem. It's about 80 per cent accurate and can give warnings 30 minutes before the irregular heartbeat starts. This model could be put on smartphones to read data from smartwatches, helping people take action to keep their heart rhythms stable. The study has been published in the journal Patterns. To develop the model, the team used 24-hour recordings collected from 350 patients at Tongji Hospital in Wuhan, China. Named WARN (Warning of Atrial fibRillatioN), the model relies on deep-learning, a form of machine-learning AI algorithms that discern patterns from historical data to forecast outcomes. Deep-learning is distinguished by its multi-layered decision-making process. Understanding this technology The researchers observed that WARN issued advance alerts, typically 30 minutes ahead of atrial fibrillation onset, marking it as the initial method to offer warnings significantly prior to the event, they remarked. Also read | European Parliament passes first-of-its-kind artificial intelligence law "We used heart rate data to train a deep learning model that can recognise different phases -- (normal) sinus rhythm, pre-atrial fibrillation and atrial fibrillation -- and calculate a 'probability of danger' that the patient will have an imminent episode," Jorge Goncalves, from the Luxembourg Centre for Systems Biomedicine (LCSB), University of Luxembourg, and the study's corresponding author, said. Goncalves mentioned that as atrial fibrillation approaches, the likelihood steadily rises until it surpasses a predefined threshold, triggering an early warning. The researchers noted that due to its low computational requirements, the AI model is "well-suited for incorporation into wearable devices." Also watch | Artificial Intelligence vs Hollywood actors: AI emerges as big concern in entertainment industry "These devices can be used by patients on a daily basis, so our results open possibilities for the development of real-time monitoring and early warnings from comfortable wearable devices," study author Arthur Montanari, an LCSB researcher, said. (With inputs from agencies)

This new AI model will predict cardiac irregularity 30 minutes before its onset

Google is pausing its AI tool that creates images of people following inaccuracies in some historical depictions generated by the model, the latest hiccup in the Alphabet-owned (GOOGL.O), opens new tab company's efforts to catch up with rivals OpenAI and Microsoft

Google pauses Gemini AI model's image generation of people due to inaccuracies

ai-model

Ai Model, News on Ai Model, stories on Ai Model, articles on Ai Model, Ai Model stories at Wion, Ai Model news, Ai Model latest news

ai model