Artificial Intelligence and Machine Learning, Government, Industry-Specific
Announcement comes hours after President Trump blacklists humanity
David Perera (@daveperera) •
February 27, 2026
OpenAI announced late Friday night that it has reached an agreement with the U.S. Department of Defense to deploy its large-scale language models to sensitive military networks.
See also: Emerging Trends in Federal Cybersecurity: Streamlining Efficiency with a Holistic IT Approach
The announcement, made by company CEO Sam Altman on social media network X, came hours after President Donald Trump directed federal agencies to stop using AI developed by OpenAI rival Anthropic. Secretary of Defense Pete Hegseth has designated Anthropic a “supply chain risk” and said he has blacklisted the company from providing business to the Department of Defense (see: Trump escalates conflict between AI and humanity).
Altman said defense officials agreed with open AI executives on the importance of “prohibiting mass domestic surveillance and human responsibility for the use of force, including autonomous weapons systems.”
“We will also build technical safeguards to ensure that our models perform as expected, which is what the U.S. Department of Transportation also wanted,” he said, referring to Defense by the Trump administration’s preferred nickname “Department of the Army.”
Altman also said he urged government officials to offer the same terms to all AI companies and “de-escalate the situation from legal and government action and move toward a reasonable agreement.”
Anthropic and the Department of Defense are locked in a disagreement over how the military should use Anthropic’s Claude models. Defense officials reportedly want broad flexibility across intelligence, cyber operations, and battlefield support functions, but Anthropic wants to maintain restrictions on domestic surveillance and autonomous weapons applications.
“We do not knowingly provide products that endanger American combatants or civilians,” Amodei said in a statement Thursday.
It is unclear how quickly OpenAI technology will be integrated into sensitive networks. Claude is the only base model approved for use in certain classified defense environments. What could accelerate that adoption is a partnership with Amazon Web Services that OpenAI announced Friday. Both companies will build runtime environments on the Amazon Bedrock platform. This platform has been recognized by the federal government as suitable for non-classified but highly sensitive data.
