OpenAI’s Human Rights Lead: What the military could do with AI ‘keeps me up at night’

Sarah Yager, OpenAI's first Human Rights and Responsible Deployment Lead, expressed concerns about the potential military applications of AI during her recent speech at the ScaleUp:AI conference in New York City. With 30 years of experience in this area, Yager emphasized the importance of addressing issues such as targeting, autonomous weapons, and the overall direction of AI in military operations.

She noted that the rapid incorporation of AI by militaries worldwide has sparked her interest in navigating these complex questions from within one of the most powerful AI companies. OpenAI contracted the Department of War (DoW) to deploy advanced AI systems in a classified environment in late February, following a public disagreement between the DoW and Anthropic over guardrails.

Yager joined OpenAI after conducting thorough due diligence, as she felt compelled to see if she could make a difference in preventing the company from appearing as a figurehead without actual influence. One of Yager's concerns includes the military's potential misuse of AI in armed conflict, particularly in targeting and the development of autonomous weapons.

She highlighted the need for improved models that adhere to international agreements, such as the laws of war established in 1949 at the Geneva Conventions. Yager also advocated for societies to rely on existing human rights frameworks, like the Geneva Conventions, as the baseline for AI development rather than leaving it to AI companies to determine what is "good."

Furthermore, she expressed reservations about discussions surrounding AI rights within Silicon Valley, feeling that these debates should focus more on human rights and the legal structures surrounding AI.