White House Rejects Anthropic’s Mythos AI Expansion
The White House has expressed opposition to a proposal by Anthropic PBC aimed at broadening access to its Mythos AI model, according to an official from the administration on Wednesday evening. Officials within President Donald Trump’s administration have communicated to Anthropic their disagreement with the company’s strategy to provide access to its Mythos technology to approximately 70 companies and organizations, as reported by a source who requested anonymity to discuss confidential discussions.
Concerns surrounding Mythos, a technology that the company claims possesses the capability to facilitate perilous cyberattacks, as previously highlighted. Concerns have been raised by US officials regarding Anthropic’s computing capacity to accommodate a growing number of Mythos users, potentially hindering the government’s effective utilization of the technology, as reported, citing anonymous sources. A White House official conveyed that the administration is navigating the delicate interplay between innovation and security, while collaborating with the private sector to guarantee the safe deployment of artificial intelligence models.
The White House has yet to provide a response to a request for comment made late Wednesday night. Anthropic refrained from providing a comment. Launched in early April, Mythos is claimed to have the capability to identify and exploit vulnerabilities across a diverse array of essential software. Anthropic has assessed the technology as too perilous for broad distribution and has consequently permitted a select group of companies to initiate testing within their own infrastructures.
However, it has reported that the model has been accessed by a limited number of unauthorized users. The risk was highlighted when a small group of unauthorized users in a private online forum accessed Mythos on the same day that Anthropic announced its limited release plan, according to a source.
