OpenAI Opens Safety Fellowship Applications to Fund External AI Safety and Alignment Researchers

OpenAI has launched a pilot OpenAI Safety Fellowship program that makes resources available to independent researchers outside the organization, with a focus on AI safety, ethics, alignment and system robustness. As frontier models rapidly become more capable, reducing the risks of loss of control, misuse and divergence from intended values has become a major issue for industry governance and technological development.

Applications for the OpenAI Safety Fellowship are now open. Selected fellows will receive a monthly stipend, research compute and mentorship, though the available information does not disclose the stipend amount, number of places or program duration. Applications close on May 3, 2026, and the program is intended to enable external researchers to pursue longer-term, independent work on AI safety and alignment.