Anthropic and the Pentagon are back at the negotiating table, FT reports

Anthropic CEO Dario Amodei returned to the negotiating table with the US Department of Defense after talks on the use of the company’s artificial intelligence tools by the military ended on Friday, The Financial Times reported.

Amodei is meeting with Emil Michael, the undersecretary of defense for research and engineering, as a last resort to reach an agreement on the terms governing the Pentagon’s access to Anthropic’s Claude models. The Times reportedquoting anonymous sources knowledgeable on the subject.

The discussions broke down Friday, with President Donald Trump ordering federal agencies to stop using Anthropic’s vehicles and Defense Secretary Pete Hegseth saying he would designate the company as a supply chain risk to national security.

Last week, Michael attacked Amodei, calling him a “liar” with a “God complex.” X post.

Acceptance of a new contract will allow the US military to continue using Anthropic’s technology. reportedly It was used in Washington’s war with Iran.

Claude became the first major model used in the government’s secret networks. $200 million contract It was awarded to Anthropic by the Department of Defense, but the company later sought guarantees that its vehicles would not be used in local surveillance or autonomous weapons. The Pentagon had requested that the army be allowed to use this technology. any legal use.

In a Friday memo seen by the FT, Amodei reportedly told staff that towards the end of negotiations with the Ministry of Defense, Anthropic had offered to accept Anthropic’s terms if they deleted “specific language relating to ‘analysis of collectively obtained data'”. He said that sentence “exactly matches this scenario that we’re most concerned about.”

Amodei also wrote in his note that the messages from the Pentagon and OpenAI, which reached a new agreement with the Department of Defense on Friday, amounted to “blatantly lying about these issues or trying to confuse them.”

The timing of OpenAI’s deal with the Pentagon, announced just hours after the White House condemned Anthropic, sparked public outrage; Claude had seen an increase in app downloads while using ChatGPT. reportedly We’ve seen app uninstalls increase.

OpenAI CEO Sam Altman later said his company “shouldn’t rush” the deal and outlined revisions to its own security measures for how the Department of Defense might use its technology.

One Publish on XAddressing the controversy further, Altman said: “In my conversations over the weekend, I reiterated that Antropik should not be designated as a group. [supply chain risk]and we hope [Department of Defense] “We offer them the same conditions we agreed on.”

Anthropic was founded in 2021 by a group of former OpenAI staff and researchers who left the firm after disagreements over its direction, and the company was marketing itself as a “security first” alternative.

Government officials have criticized Anthropic for months for allegedly being overly concerned about AI security.

A tech industry group whose members include Nvidia, Google and Anthropic sent a letter to Hegseth on Wednesday expressing concerns about his designation of a U.S. company as a supply chain risk.

The Department of Defense and Anthropic did not immediately respond to CNBC’s request for comment on the reported negotiations.