The US government granted Anthropic permission to deploy its latest model, Mythos, to a specific group of trusted partners. This move allows a limited set of US companies to access the technology before a wider release. This restrictive rollout suggests the government is managing the deployment carefully.

Semafor reports that this selective access focuses on companies deemed reliable by federal standards. By limiting the distribution of Mythos, the government maintains control over who utilizes the model's capabilities. This approach prevents a general public release while still allowing domestic firms to test the tool.

This decision reflects a strategy of gated access for high-capability AI tools. Rather than a full launch, the US is prioritizing trusted partners to ensure the model is used responsibly. This specific deployment model creates a tiered system of access for advanced AI development within the United States.