Anthropic's co-founder stated on Monday that the company is in discussions with the Trump administration regarding its advanced artificial intelligence model, Mythos. This dialogue continues despite the Pentagon having severed business ties with the company due to a contract dispute.
The disagreement between Anthropic and the Pentagon centered on guidelines for the military's use of Anthropic's AI tools. This led the Pentagon last month to designate Anthropic as a supply chain risk, prohibiting its use by the Department of Defense and its contractors.
"We have a narrow contract dispute, but I don't want that to get in the way of our deep concern for national security," said Jack Clark, a co-founder of Anthropic. "Our position is that the government must understand these things. Therefore, we are absolutely discussing 'Mythos' with them, and we will also discuss the next model with them."
The specific nature and details of Anthropic's discussions with the U.S. government, including which agencies are involved, remain unclear.
Mythos was released on April 7th and is currently Anthropic's most capable model for coding and agent-based tasks.
Experts say the model's high-level coding capabilities give it unprecedented potential for identifying cybersecurity vulnerabilities and designing methods to exploit them.
In a recent development, a federal appeals court in Washington D.C. last week declined to temporarily block the Pentagon's decision to place Anthropic on a national security blacklist. This ruling was a victory for the Trump administration, coming after a different appeals court had reached an opposite conclusion in a separate legal challenge brought by Anthropic.
Comments