Anthropic has resumed customer access to its newest AI model, Fable, following a more than two-week suspension prompted by US government security concerns. The San Francisco-based company restored access after working with the Department of Commerce to address vulnerabilities that could potentially facilitate cyber-attacks. Commerce Secretary Howard Lutnick confirmed the move, stating the government has worked closely with Anthropic to ensure alignment and strengthen national AI leadership.
Why was the Fable AI model suspended?
The US government restricted access to Fable after researchers at Amazon identified a method to bypass the model’s safety guardrails. According to the findings, the exploit allowed the AI to assist in identifying software vulnerabilities, creating a risk that malicious actors could use the technology for offensive cyber-attacks. This intervention marked a significant shift in the White House’s approach to AI regulation, moving away from a previously light touch intended to maintain a competitive edge against China.
Intelligence agencies from the Five Eyes alliance—the US, UK, Australia, Canada, and New Zealand—recently warned that frontier AI models will fundamentally transform both offensive and defensive cyber capabilities within a matter of months, not years.
How is the US government regulating AI access?
The government is now actively vetting which organizations receive access to high-powered frontier models. Anthropic has been authorized to release a related model, Mythos 5, but only to a select group of “trusted” partners for defensive cybersecurity purposes. Under the new protocols, Anthropic must proactively detect security risks, maintain standards for future model releases, and report any observed malicious activity to federal officials. Susie Wiles, Donald Trump’s chief of staff, described this collaboration as an “unprecedented” foundation for the “America First” approach to technology deployment.
What are the industry concerns regarding government oversight?
The trend toward government-led vetting has sparked debate among industry leaders. OpenAI chief executive Sam Altman expressed concern regarding the government’s role in picking specific customers for AI access. OpenAI itself recently delayed the public launch of its GPT-5.6 at the request of the US government, limiting its availability to a small cohort of vetted partners.
Comparison of Government Engagement
| Company | Regulatory Action |
|---|---|
| Anthropic | More than two-week suspension of Fable; now working with dedicated government teams. |
| OpenAI | Delayed public launch of GPT-5.6; restricted to vetted partners. |
Frequently Asked Questions
Is Anthropic’s Fable model safe to use now?
According to Anthropic, researchers from the US Department of Commerce’s Center for AI Standards and Innovation (CAISI) have tested the company’s updated safeguards and verified they are “extraordinarily strong.”
Will the government continue to control AI releases?
Yes. The government is working with the private sector to implement guardrail testing before wide-scale deployment.
How does this affect future AI development?
Companies are increasingly standing up dedicated teams to collaborate on government priorities, signaling that the future of frontier AI development will be closely tied to national security standards and federal oversight.
For developers and enterprise users, staying updated on the Department of Commerce guidelines is essential, as these standards are rapidly becoming the baseline for authorized AI usage in the United States.
Are you concerned about the impact of government oversight on AI innovation? Share your thoughts in the comments below or subscribe to our newsletter for ongoing updates on AI policy.
