A new pilot program from OpenAI funds independent safety and alignment research. The fellowship targets the next generation of technical talent to study model risks. This initiative focuses on academic rigor over product development. Researchers will gain access to resources to refine alignment techniques, providing a structured path for safety-critical experimentation.