Summaries like this, in your inbox every morning.
Sign up free →What happened
The US Department of Commerce gave Anthropic 90 minutes to address a reported 'narrow, non-universal jailbreak' in its Fable model and ordered the shutdown of both Fable 5 and Mythos 5. According to the article, Amazon CEO Andy Jassy informed the Trump administration that Fable was untrustworthy, and Anthropic and Trump administration officials are now in tense talks over whether the models could return under additional safeguards.
Why it matters
The article reports that similar jailbreaks exist in other leading models like OpenAI's GPT-5.5, yet those models have not faced comparable export controls. AI and security experts warn in an open letter that Commerce's directive 'has taken the best models away from defenders, created market uncertainty, and risked America's AI leadership without any real risk to justify it.' The timing and selective enforcement suggest this is less about genuine security than about political pressure on Anthropic, which previously refused to allow Defense Secretary Pete Hegseth to use its models to power autonomous weapons.
What to watch
Commerce indicates it might allow a narrower relaunch of the models if jailbreak issues are resolved and additional controls are put in place. The article notes that Anthropic and Trump administration officials are in tense talks over the terms of any relaunch, with the outcome uncertain given what the author describes as the administration's limited technical expertise in AI policy.
No discussion yet for this article
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
5 minutes a day. The AI essentials.
200+ sources · Email / LINE / Slack
