Anthropic and Trump administration officials continue talks over restricted access to Fable 5 and Mythos 5
The facts
- Anthropic and Trump administration officials held or continued talks on Monday to try to resolve the dispute over access to Anthropic’s latest AI models.
- No agreement had been reached by Monday in those talks.
- The U.S. government ordered Anthropic to block foreign nationals from using Fable 5 and Mythos 5, citing national security or security concerns.
- In response to the order, Anthropic disabled access to Fable 5 and Mythos 5 for all users, not only foreign nationals, in order to comply.
- The restricted models are Anthropic’s newest and most advanced systems, making the dispute consequential for the company’s customers and product rollout.
- The administration’s stated concern is that the models could be misused if safeguards were bypassed, including by foreign military or intelligence users in countries of concern.
- The move has drawn criticism from cybersecurity leaders and policy observers, who argue the restrictions could hinder cyber defense efforts and signal a case-by-case approach to AI oversight.
- It remains unresolved whether Anthropic can provide assurances that satisfy the government and allow access to the models to be restored.
Context
Why did Anthropic shut off the models for everyone instead of only foreign users?
Anthropic said it disabled Fable 5 and Mythos 5 for all users after the U.S. ordered it to block foreign nationals, because a global shutdown was the way it chose to ensure compliance with the directive Hindustan Times,Indian Express,Economic Times.
What security concern did the U.S. government cite?
Officials said they were concerned the models’ safeguards could be bypassed and that the systems could then be used in harmful ways, including by military or intelligence users in countries such as China or Russia Forbes,TimesNow,Dawn.
Why does this dispute matter beyond Anthropic?
The shutdown affects access to a leading AI company’s newest high-end models and has prompted debate over how the U.S. will regulate advanced AI tools through export controls and security directives Hill,Los Angeles Times,Verge.
Common ground, and where they split
- Common ground
- Powerful new AI models were pulled from broad use because the government treated misuse risks as serious, and both framings accept that restoring access turns on whether credible safeguards can address those security concerns without indefinite disruption.
- Where Left and Right split
- The collateral cost of a broad, ad hoc shutdown that cut off everyone, or the government’s duty to keep advanced models away from foreign military or intelligence misuse unless Anthropic can prove adequate safeguards.
How left and right read it
What stands out here is how a security order aimed at foreign nationals ended up cutting off access for everyone, including customers waiting on Anthropic’s newest systems. With no agreement yet and restoration still unresolved, the episode underscores both the real stakes of AI misuse concerns and the costs of ad hoc oversight that critics say can undercut cyber defense and create uncertainty around access.
The key facts are that the government restricted access to Anthropic’s most advanced models over stated security concerns, including potential misuse by foreign military or intelligence users, and that talks still had produced no agreement by Monday. What matters is whether Anthropic can offer assurances strong enough to restore access without brushing past a basic sovereign responsibility to keep powerful systems out of the wrong hands.
The receipts — 89 sources
Wire services (3)
Other coverage (86)
A news briefing built to show both sides.
The day’s biggest stories in one short brief, built to show both sides and the competing values behind the headlines. Free in your inbox.