OpenAI Signs AI Deal With the Pentagon

Sam Altman, the CEO of OpenAI, announced that the company has reached an agreement with the Pentagon. The deal allows OpenAI’s artificial intelligence tools to operate within the U.S. military’s classified systems.

The announcement came shortly after Altman spoke at the AI Summit in New Delhi, India, on February 19, 2026. According to Altman, the agreement includes safety limits similar to those that another AI company, Anthropic, had previously requested during its own discussions with the Pentagon.

ALSO READ: WhatsApp Plans Advertising Rollout with Optional Paid Subscription

Safety Rules and Technical Safeguards

Altman explained that two major safety principles guide the agreement. First, OpenAI’s systems cannot be used for domestic mass surveillance of U.S. citizens. Second, humans must remain responsible for decisions involving the use of force, even when autonomous weapons systems are involved.

Altman said the Department of Defense supports these principles and has already reflected them in U.S. law and policy. He added that OpenAI will also implement technical safeguards to ensure its AI models follow these rules.

To support this effort, OpenAI plans to send engineers to work directly with the Pentagon. These engineers will monitor the systems and help ensure the models operate safely and responsibly.

Tensions With Anthropic and Government Response

The OpenAI agreement was announced on the same day that U.S. President Donald Trump ordered federal agencies to stop using AI tools developed by Anthropic. Meanwhile, Defense Secretary Pete Hegseth labeled Anthropic a “supply chain risk.”

The dispute began when Anthropic refused to accept certain Pentagon conditions regarding the use of its AI in autonomous weapons and mass surveillance programs. In response, the company announced plans to challenge the designation in court. Normally, such a label is applied to companies linked to foreign adversaries.

It remains unclear how OpenAI’s agreement differs from the conditions Anthropic rejected. Both the Pentagon and OpenAI have been asked to clarify the details of the deal.

Meanwhile, Pentagon technology chief Emil Michael supported the partnership. He stated that reliable cooperation with technology companies is critical as the military enters what he called the AI Age, especially when decisions may affect the lives of soldiers.