OpenAI Amends Pentagon Deal Amid Surveillance Concerns
OpenAI is revising its recently established contract with the U.S. Department of Defense to explicitly prohibit the use of its artificial intelligence systems for domestic surveillance, following public outcry and concerns raised by employees. The move comes after a deal was struck allowing OpenAI’s AI models to operate on classified military networks, a contract that followed the Trump administration’s decision to restrict the use of rival Anthropic’s AI tools.
Deal Sparks Controversy
The initial agreement between OpenAI and the Pentagon, announced on Friday, February 28, 2026, quickly drew criticism. Concerns centered on the potential for the technology to be used for mass surveillance of U.S. Citizens or in the development of autonomous weapons systems. These concerns were amplified by the context of the deal being announced just hours after President Donald Trump directed federal agencies to stop using Anthropic’s Claude system, and hours before military strikes were carried out in Iran.
Altman Addresses Concerns
OpenAI CEO Sam Altman acknowledged the rushed nature of the agreement and admitted a mistake in its initial rollout. In a post on X (formerly Twitter) on Monday, March 3, 2026, Altman shared an internal memo outlining the amendments. “We shouldn’t have rushed,” Altman stated, adding that the company was “genuinely trying to de-escalate things and avoid a much worse outcome, but I think it just looked opportunistic and sloppy.”
The amended contract will include language stating that, “Consistent with applicable laws, including the Fourth Amendment to the United States Constitution, National Security Act of 1947, FISA Act of 1978, the AI system shall not be intentionally used for domestic surveillance of US persons and nationals.” Altman also affirmed that the Department of Defense has stated OpenAI’s tools will not be used by intelligence agencies like the NSA, though any such use would require a further contract modification.
Anthropic’s Standoff with the Pentagon
The OpenAI deal followed a breakdown in negotiations between the Pentagon and Anthropic. The Department of Defense had sought to allow AI models to be used for “all lawful purposes,” while Anthropic sought explicit prohibitions on domestic surveillance and autonomous weapons. President Trump subsequently criticized Anthropic and designated the company as a “supply-chain risk,” directing federal agencies to phase out its products. More than 60 OpenAI and 300 Google employees signed a letter supporting Anthropic’s position.
Employee and Public Reaction
The OpenAI-Pentagon deal triggered protests in San Francisco and London, and a boycott campaign launched by the advocacy group QuitGPT. The decision also sparked internal debate within OpenAI, with many employees voicing support for Anthropic’s stance.
Looking Ahead
OpenAI’s revisions aim to address immediate concerns, but the debate surrounding the ethical implications of AI in military applications is likely to continue. The company has committed to working with the Pentagon on technical safeguards and acknowledging the complexities and uncertainties surrounding the technology. The situation highlights the growing tension between national security interests and the need for responsible AI development.