US Govt Reviews OpenAI & Anthropic AI Models: De-Facto Licensing

The US government is directly reviewing new AI model launches from companies like OpenAI and Anthropic. This creates a de-facto pre-approval system for cutting-edge artificial intelligence. OpenAI adjusted its GPT-5.6 model after government consultations, which included reviews of cybersecurity and safety features. The government also requested limitations on its release. Anthropic faced even stronger intervention. The government issued an export control directive, temporarily suspending foreign access to its 'Claude Mythos5' and 'Claude Fable5' models. While the export control was lifted for Fable5, Mythos remains restricted for foreign use. Anthropic now provides pre-release access to models related to national security. AI companies are also strengthening user identity verification. Anthropic, for example, now requires users to provide a passport or driver's license in certain situations, along with a live selfie for verification. OpenAI introduced a similar age verification process earlier this year. This means that new AI models are now subject to significant government oversight before they reach the public.