OpenAI Human Rights Lead Warns of Military AI Risks

Summary

Sarah Yager, OpenAI’s first human rights and responsible deployment lead, says military use of AI, especially for targeting and autonomous weapons, is a major concern. She took the role in June after decades of work on military applications and after OpenAI secured a contract with the Department of War to deploy advanced AI systems in classified environments. The contract followed a dispute with Anthropic, whose proposed restrictions included bans on mass public surveillance and fully autonomous weapons; OpenAI said it obtained similar guardrails. Yager initially worried that OpenAI might use her as a branding exercise, but interviews with employees persuaded her to join and try to influence deployment from inside the company. She says she has worked with researchers and engineers to improve models’ responses to international humanitarian law, including the laws of war established through the 1949 Geneva Conventions. International experts tested the models by entering prompts for high-stakes military scenarios, after which engineers adjusted them. Yager argues that human rights should be the baseline for AI governance rather than values defined solely by technology companies. She distinguishes debates about whether AI agents should receive legal, labor, or tax status from human rights, and says affected communities need a larger role in AI decisions. She is also helping the Government Accountability Office and an Office of Inspector General develop tools to monitor government AI use, which she considers important for democratic accountability.