OpenAI Is Paying $445K to Study Self-Improving AI Risks

OpenAI has posted a job listing for a Senior Machine Learning Engineer on its Preparedness team, offering up to $445,000 in total compensation. The role is focused on studying risks from self-improving AI systems, according to the job description. The listing signals growing industry concern about AI control risks, where an AI system could recursively improve its own capabilities beyond human oversight.

The Preparedness team is responsible for evaluating and mitigating risks from frontier AI models, including those related to self-improvement. The job description states that the engineer will work on 'understanding and mitigating risks from self-improving AI systems,' a topic that has gained attention as models become more autonomous.

The listing comes amid broader discussions about AI safety and governance. OpenAI has previously published research on risks from advanced AI, including a report on potential control problems. The company has also faced criticism from safety researchers who argue that labs are not moving quickly enough to address these risks. Industry observers note that the salary reflects the high demand for AI safety talent, as companies compete for experienced researchers.