OpenAI Reaches Agreement With Pentagon for Classified AI Work

OpenAI has signed an agreement with the Department of Defense to provide artificial intelligence systems for classified military applications, marking a significant shift in the company's defense policy.

The deal, announced Friday, allows OpenAI technologies to be used for any lawful defense purpose, including classified environments. The company stated it found ways to maintain its safety principles while meeting Pentagon requirements.

The agreement comes amid tensions between tech companies and the Pentagon over military AI use. Anthropic, another major AI company, reportedly clashed with defense officials over similar terms earlier this week, with Anthropic calling Pentagon blacklisting of its technology "legally unsound."

What the Agreement Covers

Under the terms, OpenAI models can be deployed in classified settings but will not be used for domestic surveillance or autonomous offensive weapons, according to statements from the company. Sam Altman, OpenAI CEO, held an AMA session on Saturday to address questions about the partnership.

The Pentagon has been actively seeking AI partnerships as part of broader efforts to modernize military capabilities. The defense sector has emerged as a major market for AI companies, though the partnerships have drawn criticism from researchers and advocacy groups concerned about military applications of the technology.

Background

OpenAI updated its usage policies last year to allow military applications, reversing an earlier prohibition. The company has faced pressure to define clear boundaries for defense work while competing for government contracts.

Industry analysts estimate the defense AI market could reach $30 billion annually by 2030, making Pentagon partnerships increasingly important for AI companies seeking growth beyond consumer and enterprise markets.