The Trump administration's approach to AI has been erratic, despite President Trump's stated goal to accelerate the AI industry and beat China. Last week, Anthropic released Fable 5, an advanced AI system, which is an updated public version of Claude Mythos Preview. Mythos Preview was initially restricted to cybersecurity partners due to its hacking capabilities, but Anthropic developed guardrails and conducted third-party testing, including with the U.S.
government, before Fable 5's release. Cybersecurity experts even found Fable 5 too restrictive. On Friday, the White House deemed Fable 5 a national security threat, reportedly giving Anthropic 90 minutes to remove Fable 5 and Mythos 5. When Anthropic did not comply, the government issued an export control, preventing any foreign national, even Anthropic employees in the U.S., from using the models. Consequently, Anthropic shut down the bots for all customers, meaning American companies and the U.S. government cannot use what is potentially the world's most powerful AI, for unclear reasons.
This is an editorial summary for informational purposes. All rights to the original content belong to the respective author and publication.
No comments:
Post a Comment