Anthropic president Daniela Amodei has questioned OpenAI’s Pentagon agreement. Speaking alongside her brother, Anthropic CEO Dario Amodei, she suggested the rival company may have acted opportunistically when it reached a deal after Anthropic disputed the US government’s military use of AI. Her comments came during an interview with Walter Isaacson for his podcast “The Builders,” according to an excerpt the New York Post obtained.“Was this a really principled stand that another technology company took based on their own ethics, or was this an opportunistic moment?” Daniela said when asked about OpenAI stepping in. The criticism centers on whether OpenAI’s agreement preserved meaningful safeguards after Anthropic refused to remove restrictions on mass domestic surveillance and fully autonomous lethal weapons.
Why Anthropic refused the Pentagon ’s request
The dispute began in February, when Anthropic declined to abandon those restrictions on its AI models. Pentagon officials argued that defense contractors should allow their technology for “all lawful uses” and that private companies should not determine military operations.US President Donald Trump subsequently directed federal agencies to stop using Anthropic’s products. The Pentagon also designated the company a supply-chain risk and restricted its role in defense work.OpenAI later reached its own agreement with the Pentagon. The company has maintained that the deal includes explicit restrictions against mass domestic surveillance, directing autonomous weapons systems and making high-stakes automated decisions.Daniela’s remarks questioned the circumstances behind that agreement. However, the report didn’t confirm that OpenAI abandoned the safeguards it says are included in its deal.
Pentagon dispute widened rift between Anthropic and OpenAI
The disagreement follows years of tension between the companies. Dario and Daniela left OpenAI in 2021 and founded Anthropic amid differences over AI safety, corporate governance and OpenAI’s increasingly commercial direction.During the Pentagon dispute in March, Dario Amodei reportedly described OpenAI’s account of its military agreement as “straight up lies” and “safety theater” in a leaked internal message. He later apologized for his tone while maintaining Anthropic’s position.In the Isaacson interview, Dario Amodei also raised concerns about concentrating control over AI-powered weapons. He warned that the prospect of “one person having their finger on a button that controls a million robots or a million drones” was “kind of fundamentally undermining to democracy”.
Anthropic continues its legal challenge
Anthropic challenged the administration’s actions in US court. According to the report, it secured a favourable ruling in California in August, before a separate federal appeals court in Washington upheld a Pentagon exclusion the following month.The company sought a rehearing of the Washington decision earlier this month. Its prospectus also identified the confrontation as a business risk, warning that the litigation could affect sales to government agencies and federal contractors.
