Sam Altman is positioning ChatGPT creator OpenAI as a potential intermediary in an ongoing debate over how the Pentagon should deploy advanced artificial intelligence.
In a memo to staff on Thursday, viewed by the Wall Street Journal, Altman said OpenAI is in discussions with the Defense Department about using its AI models in classified environments. He emphasized that any agreement would require firm limits on applications such as domestic surveillance and autonomous offensive weapons. No deal has been finalized, and a person familiar with the talks said negotiations could still fall through.
Altman framed the disagreement between OpenAI and rival AI company Anthropic, which develops the Claude chatbot, as less about specific use cases and more about decision-making authority. While expressing general support for Anthropic’s proposed “red lines,” Altman said private companies should not override elected government officials on matters of national security.
“For all the differences I have with Anthropic, I mostly trust them as a company, and I think they really do care about safety,” Altman told CNBC on Friday, according to the AP. “I’ve been happy that they’ve been supporting our warfighters.” He added, “I’m not sure where this is going to go.”
According to the Journal, OpenAI has proposed technical safeguards—such as hosting models in the cloud and embedding company staff alongside government users—to help enforce usage restrictions and potentially establish a framework other AI developers could follow.
Anthropic CEO Dario Amodei previously disclosed that his company rejected Pentagon requests for “all lawful uses” of its technology. A senior Defense Department official responded that the agency would not allow technology companies to determine the scope of Americans’ civil liberties.
Meanwhile, Elon Musk’s AI venture xAI agreed to the Defense Department’s “all lawful uses” condition. However, Axios reported that xAI’s Grok chatbot is not widely viewed as a full-scale alternative to Claude.
Some observers have criticized how the Pentagon is handling the broader issue. Retired Air Force Lt. Gen. Jack Shanahan wrote on social media, “Painting a bullseye on Anthropic garners spicy headlines, but everyone loses in the end.”