04/03/2026
On March 3, 2026, OpenAI, led by Sam Altman, amended its recent agreement with the U.S. Department of Defense (Pentagon) after facing significant public backlash. The company has now explicitly prohibited the use of its AI systems for spying on American citizens.
The move comes shortly after reports of a sharp rise in ChatGPT uninstalls, with some coverage citing a 295% increase following news of the military partnership. The incident highlights growing public sensitivity around AI companies working with defense agencies, particularly on issues related to privacy, surveillance, and national security.
At the same time, the development reflects wider tensions in the AI industry. OpenAI’s deeper engagement with the Pentagon follows challenges faced by Anthropic, which reportedly encountered difficulties in defense-related collaborations, including being labeled a “supply chain risk” in certain contexts. Despite restrictions, reports suggest its models were still used in limited scenarios, fueling further debate.
These dynamics have also impacted user behavior. ChatGPT has faced user churn, while Claude - Anthropic’s AI assistant - has reportedly gained popularity as some users shift platforms over ethical and trust concerns.
This situation underscores a critical moment for AI companies: balancing innovation, government partnerships, national security interests, and public trust. As AI systems become more deeply embedded in defense and infrastructure, transparency and clear usage boundaries are becoming non-negotiable.