Two weeks after the Trump administration mandated Anthropic suspend foreign access to its flagship Mythos AI models, negotiations remain stalled with no clear end in sight, leaving the company’s most powerful systems offline and casting shadow over US AI export regulations.
- Anthropic forced to suspend Mythos access by export controls targeting foreign nationals
- Negotiations with Trump administration extend beyond two weeks without resolution
- Regulatory uncertainty clouds future of AI export controls and US industry growth
What happened
On June 12, the Trump administration issued an export control order requiring Anthropic to suspend access to its Mythos 5 and Fable 5 AI models for any foreign nationals, including those employed by Anthropic. This directive came amid security concerns related to the models’ capabilities and potential misuse. As a result, Anthropic promptly took its Mythos-class models offline to comply with the mandate.
Since then, Anthropic has engaged in intensive negotiations with government officials, dispatching executives to Washington, DC to seek a resolution. However, no official updates or progress announcements have emerged, and the company has declined to comment publicly. The negotiations are reportedly complicated by the lack of an established regulatory framework suited to AI systems that exhibit dual-use technology characteristics.
Why it matters
The extended suspension of Mythos-class AI models poses serious commercial challenges for Anthropic, which had positioned these advanced models as key revenue drivers ahead of an anticipated IPO. Mythos-class models command higher input token prices due to their superior performance, and their downtime directly impacts Anthropic’s financial outlook and market position.
Beyond Anthropic, the case highlights broader issues surrounding the US government’s approach to AI export controls. Traditional export control regimes rely on assessing physical products during production, whereas AI models like Mythos present unique regulatory challenges due to their software-based, rapidly evolving nature. This regulatory uncertainty could hinder innovation and investments across the AI sector.
What to watch next
Attention will focus on the outcome of negotiations between Anthropic and the Trump administration, including whether an agreement can be reached that allows Mythos models to be reinstated with appropriate safeguards. The government’s stance on expanding export controls to other AI companies and technologies will also be critical to monitor, as it will signal future regulatory directions for AI exports.
Industry observers and security experts will be watching for any new frameworks or policies designed to balance national security concerns with innovation needs. The evolution of AI export regulations could set precedents for dealing with software-based dual-use technologies and influence global competition in AI development.