The US government has no intention of expanding export controls against Anthropic to other artificial intelligence companies.

A government official in the United States revealed on Saturday that the White House will not extend export controls against Anthropic's high-end AI models to other artificial intelligence companies. The official stated that the reason for this control was Anthropic's refusal to fix security vulnerabilities in its recent releases of the Fable 5 and Claude Mythos 5 models. Anthropic, however, objected, stating that AI models developed by other companies also have similar risks. In a blog post released on Friday, the company stated that several publicly available AI models have been found to have similar vulnerabilities. For example, OpenAI's latest model, GPT-5.5, released in April this year, performed similarly to the Mythos model on some cybersecurity testing metrics. The official explained that the US is concerned that the Fable model could still be used to detect vulnerabilities in sensitive systems and is also worried that advanced AI technology could fall into the hands of foreign hostile forces.