OpenAI CEO Admits Pentagon Deal Was 'Rushed and Sloppy'

Sam Altman says the company is working to clarify principles around AI use after backlash from rival Anthropic.

Published on Mar. 4, 2026

Got story updates? Submit your updates here. ›

OpenAI CEO Sam Altman acknowledged that the company's recent deal to allow the Pentagon to deploy its AI models was 'opportunistic and sloppy.' Altman said OpenAI is working to add language to the agreement to ensure its technology is not used for domestic surveillance of Americans or by intelligence agencies like the NSA. The admission comes after rival Anthropic, which had demanded its own AI not be used for mass surveillance or autonomous weapons, was blacklisted by the Defense Department.

Why it matters

The clash between OpenAI and Anthropic over Pentagon contracts highlights the growing tensions in the AI industry around the use of powerful language models for military and intelligence purposes. OpenAI's rush to secure a deal with the Defense Department after Anthropic's standoff raised concerns about the company's principles and commitment to ethical AI development.

The details

OpenAI CEO Sam Altman said the company's agreement to let the Pentagon deploy its AI models on its classified network was hastily put together. Altman said OpenAI is now working to 'make our principles very clear' in the deal, including ensuring the technology is not used for domestic surveillance of Americans or by intelligence agencies. The admission comes after Anthropic, a rival AI company, had demanded its own products not be used for mass surveillance or autonomous weapons, leading to the Pentagon blacklisting Anthropic. OpenAI announced its Pentagon deal just hours after Anthropic's blacklisting.

  • On March 4, 2026, OpenAI CEO Sam Altman acknowledged the company's Pentagon deal was 'rushed and sloppy'.
  • On the same day, Altman said OpenAI is working to add language to the agreement to clarify the principles around AI use.

The players

Sam Altman

The CEO of OpenAI, an artificial intelligence research company.

Anthropic

An AI company that had demanded its technology not be used for mass surveillance of Americans or autonomous weapons deployment, leading to a clash with the Pentagon.

Department of Defense

The U.S. government agency that reached a deal with OpenAI to deploy the company's AI models on its classified network, shortly after blacklisting rival Anthropic.

Got photos? Submit your photos here. ›

What they’re saying

“We shouldn't have rushed to get this out on Friday. The issues are super complex, and demand clear communication.”

— Sam Altman, CEO, OpenAI (latimes.com)

“No amount of intimidation or punishment from the Department of War will change our position.”

— Dario Amodei, Co-founder, Anthropic (latimes.com)

What’s next

OpenAI plans to host an all-hands meeting this week to address employees' questions about the Pentagon deal.

The takeaway

The clash between OpenAI and Anthropic over Pentagon contracts highlights the growing tensions in the AI industry around the ethical use of powerful language models for military and intelligence purposes. OpenAI's rush to secure a deal with the Defense Department after Anthropic's standoff raised concerns about the company's principles and commitment to responsible AI development.