san francisco: Anthropic announced on June 26 that the U.S. government has granted permission for a small group of U.S. cybersecurity companies to access its powerful AI model, Mythos 5.
Access to Mythos 5 had been blocked by US authorities due to national security concerns.
Anthropic has had a rocky relationship with the Trump administration in recent months, but a company spokesperson said it will continue discussions with the government to “expand access to Mythos 5 and make Fable 5 available to the general public again.”
On June 12, the government abruptly forced Anthropic to cut off access to two of the company’s most advanced artificial intelligence models after discovering vulnerabilities in security measures put in place to prevent misuse of the tools.
The strong action against Anthropic prompted accusations of government overreach.
In previous clashes with the White House, Anthropic infuriated President Donald Trump’s team by refusing to allow its technology to be used for mass surveillance or autonomous weapons, and the Pentagon terminated its contract with the company.
Early on the morning of June 26, Anthropic’s main rival, OpenAI, announced a new model, GPT-5.6, with limited access and government validation on a client-by-client basis.
Under pressure over the novelty of its capabilities, President Trump signed an executive order in June that established a voluntary federal review of national security risks before releasing advanced AI models.
The White House has provided little clarity on how it will enforce the executive order in which companies are considered voluntary participants or what models will be subject to its vetting rules. –AFP
