Inspired by this story? Turn the idea into your own AI art in seconds — free to start, no card required.
Anthropic's Mythos 5 model is back online for over 100 companies and government agencies after a two-week negotiation with the Trump administration that ended the AI provider's most significant service disruption to date.
• Over 100 companies and government agencies now have authorized access to Mythos 5, including their non-American employees • The consumer-facing Fable 5 model remains offline with no announced timeline for restoration • Anthropic's enterprise customers can resume AI-powered workflows that were halted during the two-week blackout • The selective restoration suggests the Trump administration achieved its goal of tighter controls over advanced AI model distribution • Image generation workflows using Mythos-powered tools can now resume for authorized enterprise users
The Mythos 5 restoration covers a substantial user base across corporate and government sectors, according to a government letter to Anthropic viewed by The Verge. The authorization extends to non-American employees at these organizations, indicating the deal addresses international access concerns that likely complicated the original negotiations.
For AI art creators working within these authorized organizations, the restoration means access to one of the most capable multimodal models returns after a disruptive two-week gap. Mythos 5's image understanding and generation capabilities had become integral to many enterprise creative workflows before the sudden shutdown.
While enterprise users regain access, Anthropic's consumer-facing Fable 5 model stays offline with no announced restoration timeline. This creates a two-tier system where corporate users can access Anthropic's most advanced capabilities while individual creators remain cut off from the public version.
The ongoing Fable 5 blackout leaves individual AI artists and smaller studios without access to Anthropic's flagship image generation capabilities, potentially driving them toward alternatives like OpenAI's GPT-5.6 models or other providers that avoided similar regulatory pressure.
The selective restoration reflects the Trump administration's success in implementing stricter controls over advanced AI model distribution. Rather than a complete ban, the government appears to have negotiated a framework that maintains enterprise access while restricting broader public availability.
This approach mirrors the administration's handling of OpenAI's GPT-5.6 release, which was also delayed and restructured following government security concerns. The pattern suggests AI providers now face routine government oversight of major model launches, with potential impacts on release timelines and access tiers for creative tools.
The Mythos crisis marks a turning point where government intervention directly shapes which AI capabilities reach creators and when. For the AI art community, this precedent raises questions about future model availability and whether similar disruptions could affect other providers' image generation tools.
