OpenAI Faces Backlash After Pentagon AI Deal, Pauses Deployment to US Intelligence Agencies
OpenAI is facing mounting criticism after its recent artificial intelligence partnership with the US Department of Defense triggered internal dissent and a widespread online backlash. The controversy has led the company to temporarily halt the deployment of its AI technology to US intelligence agencies, including the National Security Agency (NSA).
The debate began when OpenAI announced a defence-related agreement with the Pentagon. CEO Sam Altman had earlier suggested that if the collaboration succeeded, the company might ultimately be viewed as having helped ease tensions between the US government and the AI industry, particularly after rival AI firm Anthropic declined to sign a similar deal over ethical concerns.
However, Altman also warned that the arrangement could appear “rushed, sloppy and opportunistic” from the outside. Those concerns appear to have materialised as criticism quickly grew both inside and outside the company.
During an internal all-hands meeting, Altman reportedly clarified that OpenAI employees would not have authority over how the US military ultimately uses the company’s technology. According to a Bloomberg report, he told staff, “You do not get to make operational decisions.”
Anthropic had reportedly sought similar assurances before entering negotiations with the Pentagon. The company eventually stepped away from the agreement, citing fears that its AI models could be used for mass domestic surveillance or autonomous weapons systems.
Altman addressed such concerns while speaking about past uses of AI in military operations. As quoted in reports, he said, “So maybe you think the Iran strike was good, and the Venezuela invasion was badYou don’t get to weigh in on that.” His comments highlighted the limits of OpenAI’s influence over government decisions once technology is deployed.
Despite the criticism, both the Pentagon and OpenAI maintain that the AI systems involved in the agreement will only be used for “all lawful purposes.” Altman has also stated that the contract has since been revised to ensure the company’s ethical boundaries are respected.
The deal has nevertheless triggered significant public reaction. Social media campaigns calling for a boycott of ChatGPT gained traction, reportedly causing uninstallations of the app to surge by nearly 300 percent. Reports suggest that the company lost more than 1.5 million subscribers within 48 hours of the announcement, with many users shifting to Anthropic’s Claude chatbot.
Internal unease has also surfaced. Earlier, around 100 OpenAI employees joined 871 Google employees in signing a statement supporting Anthropic’s position and opposing the use of AI for military surveillance.
The situation intensified when OpenAI’s Vice President of Research, Max Schwarzer, left the company to join Anthropic. While he did not directly link his departure to the Pentagon deal, he praised the rival firm’s approach, writing on X, “I have also been very impressed with Anthropic's talent, research taste and values.”
Amid the growing controversy, OpenAI confirmed that it would pause the deployment of its AI systems to certain defence intelligence agencies. OpenAI researcher Noam Brown clarified on X, “OpenAI will not be deploying to the NSA or other DoW intelligence agencies for now.”
Brown acknowledged that the agreement had been rushed and that some concerns remained unresolved, particularly regarding “some novel ways that AI could potentially enable legal surveillance.” He added that OpenAI intends to address these issues through democratic and regulatory processes before moving forward.
The dispute has also sparked tensions between leading AI companies. Anthropic CEO Dario Amodei criticised OpenAI’s defence of the deal, claiming the arguments were misleading. In an internal memo, he said the public could see through the “gaslighting” and noted that many were now viewing Anthropic as “heroes” for refusing the Pentagon’s demands.