Public Backlash Surges Against OpenAI's Pentagon Deal
OpenAI's partnership with the DoD is facing significant public backlash, with U.S. uninstalls of the ChatGPT app spiking 295% and 1-star reviews up 775% post-announcement. Amid the fallout, OpenAI has updated its agreement with the DoD to explicitly clarify bans on domestic surveillance and autonomous weapons.
The controversy erupted after the Pentagon terminated a potential $200 million contract with AI company Anthropic. Anthropic refused to remove safeguards that would prevent its model, Claude, from being used for mass domestic surveillance or in fully autonomous weapons systems, leading the Pentagon to designate the company a "supply chain risk." Hours after the Anthropic deal fell apart, OpenAI, which had quietly removed its ban on military applications in 2024, secured the partnership. CEO Sam Altman later admitted the deal was "rushed" and that the "optics don't look good," acknowledging the rollout appeared "opportunistic and sloppy." The agreement is structured as a $200 million "other transaction agreement" (OTA), a flexible mechanism designed to speed up prototyping of advanced AI for challenges in "warfighting and enterprise domains." The work, managed by the Pentagon's Chief Digital and Artificial Intelligence Office (CDAO), is set for completion by July 2026. OpenAI asserts its agreement has stronger guardrails than previous deals, including Anthropic's. The company established "red lines" prohibiting use for mass domestic surveillance or to direct autonomous weapons, and added a third ban on high-stakes automated decisions like social credit systems. As a technical safeguard, the models will only be deployed via the cloud, not on "edge devices" like drones. Critics highlight that the contract allows the AI to be used for "all lawful purposes"—language Anthropic rejected—and that the prohibitions are tied to current laws like the Fourth Amendment and DoD Directive 3000.09. While OpenAI says the contract freezes the rules to today's laws, the language leaves room for interpretation should those laws or policies change. The partnership falls under the DoD's broader push for AI adoption, guided by its Data, Analytics, and Artificial Intelligence Adoption Strategy. Government contractors working on AI projects are expected to align with the DoD's five ethical principles: that AI be Responsible, Equitable, Traceable, Reliable, and Governable.
