Anthropic CEO Amodei a 'weirdo', says White House, as Tom Brown replaces him for crucial talks

White House described Anthropic CEO Dario Amodei as a "weirdo" during discussions held over the export ban on “Mythos” and “Fable” bots, according to a report from Wired. He has been replaced by co-founder Tom Brown in the crucial talks, with Sarah Heck, the company’s public policy chief, also present. "Tom Brown is not being a weirdo like Dario and can actually engage," the source told the outlet. The Trump administration slapped strict foreign export controls over Anthropic regarding its new AI systems “Mythos” and “Fable", citing national security issues. The order came just three days after the bots were launched. Anthropic pulled the models entirely on June 12.

According to the report, officials were having a hard time talking to Amodei, who was being difficult and didn’t listen to the concerns. Sources told Wired that Anthropic and the White House have had multiple calls over the export ban, and they have been going well ever since Amodei was replaced.

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

Anthropic's latest AI bots exposed sensitivities in critical systems

"Mythos” and “Fable" faced immediate foreign blocking after Amazon CEO Andy Jassy warned that they have evidence that their safety guardrails can be bypassed. This happened after Anthropic partnered with major firms to test Mythos and Fable under Project Glasswing, following claims that they were too powerful and could trigger doomsday. Reports suggest that the National Security Agency found during testing that Mythos identified vulnerabilities in highly secure government systems within hours.

Also Read: Company accidentally blows away Rs 4,800 crore on Claude AI in a month because of a tiny oversight

Even though the talks are going well, it is unclear whether the export ban would be lifted. The White House is particularly seeking built-in guardrails in Mythos and Fable that cannot be jailbroken, amid concerns that they can facilitate complex cyberattacks. Amodei flew to Washington DC to alleviate all concerns, saying they were overblown and hindered critical cybersecurity defence efforts. The company argues that just because the bots flagged the sensitivities in the system did not mean they could exploit them.

European Union raises concerns about ban on Anthropic's AI systems

At the G7 Summit, the European Union raised concerns about the ban and called for a "trusted partners" plan, as part of which selective access is granted to advanced AI models from US firms such as Anthropic. The diplomats argued that certain companies and countries should be given access to American AI tools not open to export to strengthen cybersecurity defences against rivals.

The feud between the White House and Anthropic started after Amodei refused the Pentagon's request for unchecked access to Anthropic's AI bots during contract negotiations. The company demanded limits on their use in mass surveillance and weaponry.