Anthropic pulls Fable 5 and Mythos 5 offline under export order
The San Francisco AI company says it shut off both models Friday to follow a new federal directive. Fresno-area teams using Fable this week may see errors.
Anthropic pulls Fable 5 and Mythos 5 offline under export order
Key Takeaways
- Anthropic took Fable 5 and Mythos 5 offline on Friday, June 12.
- The company says it acted to comply with a new federal export-control directive.
- Fable 5 launched widely this week; Mythos 5 access was already restricted.
- Fresno-area developers who opted into Fable 5 could see temporary disruptions.
The switch flipped Friday evening. Anthropic said it pulled access to its newest artificial intelligence models, Fable 5 and Mythos 5, to follow a fresh order from Washington. For Central Valley users, this reads like a heads-up more than a headline, since any Fresno or Modesto shop that tried Fable this week could run into errors this morning.
What changed Friday
In a statement, the San Francisco company said it received a government directive Friday afternoon and took both models offline to keep them from being used by foreign nationals. The move follows a week in which Fable 5 rolled out broadly. Mythos 5, the more powerful model, had already been held back over cybersecurity concerns.
Anthropic said it disagrees with how the directive came down and that the order did not spell out the national security issue at stake. The company called the situation a misunderstanding and said it wants a clear, public process for stopping unsafe deployments. It also said it hopes to restore access as soon as possible.
What Central Valley users should expect
If your team in Fresno County or Merced turned on Fable 5 this week for coding help or customer-support pilots, expect timeouts and fallback behavior until Anthropic flips the models back. Vendors that route work to Anthropic’s newest engines could see degraded features. Existing tools tied to older versions should keep running, which matters for small agencies and ag-tech shops that haven’t changed defaults yet. For now.
One developer at a downtown Fresno co-working space pointed at a frozen prompt window, a lukewarm can of Sprite under the monitor.
What officials and the company said
The company’s statement framed the shutdown as compliance, not choice. It argued the government should have the ability to block unsafe AI, but only through a transparent and technical process set in law. The Commerce Department did not respond to a request for comment Friday night.
The action lands 10 days after President Donald Trump signed an executive order laying out a voluntary federal review for high-end AI systems before public release. That timing will matter for Valley schools, city IT offices, and growers’ software vendors that are weighing which models to trust in procurement this summer.
"We believe the government should have the ability to block unsafe deployments... This action does not adhere to those principles," Anthropic said. It added that it plans to restore access "as soon as possible."
Central Valley AI is produced by the CVAI Newsdesk team and developed by Kaweah Tech, a regional firm that builds, deploys, and integrates AI solutions for businesses across California's Central Valley.