OpenAI on Wednesday announced a pilot program to fund independent researchers working on AI safety and alignment. The OpenAI Safety Fellowship will provide financial support and resources to individuals outside the company’s direct research teams. The goal is to develop new talent and broaden the field of study beyond existing corporate structures.
The program will select up to ten fellows for a six-month period. Each will receive a stipend to cover living costs and research expenses. Applicants must propose projects focused on safety risks or alignment challenges in large language models. OpenAI stated the fellowship is not tied to proprietary projects or commercial outcomes.
This initiative follows criticism that AI development is concentrated in a few corporate labs. Independent researchers often lack funding and institutional support. The fellowship aims to address this gap by providing structured opportunities for external contributors.
OpenAI’s vice president of safety, Sasha Luccioni, will oversee the program. She emphasized the need for diverse perspectives in safety research. The selection process will prioritize applicants from underrepresented backgrounds and regions.
Applications open next month. The first cohort of fellows will begin work in early 2025. OpenAI has not disclosed the total budget for the program.
Source: openai.com