OpenAI has announced that it will amend its contract with the US Pentagon, a move that comes in the wake of public backlash over the company's AI technology potentially being used for mass surveillance. CEO Sam Altman posted an internal memo on the
Altman wrote in the memo: "In accordance with applicable laws, including the Fourth Amendment to the U.S. Constitution, the National Security Act of 1947, FISA 1978, and other applicable laws, AI systems may not be intentionally used for domestic surveillance of U.S. personnel and nationals." He also added that the Pentagon confirmed that its services will not be used by the Department of War intelligence agencies (such as the National Security Agency), and such services require subsequent modifications to the contract.
The decision, which comes as OpenAI signed a deal with the Pentagon last Friday to deploy its AI models on classified military networks, comes as talks with Anthropic broke down, with the latter rejecting military use and drawing red lines, including a ban on large-scale domestic surveillance and fully autonomous weapons. Altman acknowledged that the company had "done it wrong" in its "rush" to move forward with the agreement, saying the issue was complex and required clearer communication and said the move was intended to "de-escalate the situation and avoid a worse outcome."
The announcement of the agreement quickly triggered a backlash, with nearly 500 OpenAI and Google employees signing an open letter supporting Anthropic's position; protests broke out outside the OpenAI offices in San Francisco and London, and the opposition group QuitGPT launched a boycott and planned further demonstrations on Tuesday. Anthropic has not yet responded to a request for comment.