Two years of safety researcher departures have prompted OpenAI to invest outside its own walls. On April 6, OpenAI announced a Safety Fellowship, a pilot program offering independent researchers funding and support to work on AI safety and alignment.
Alignment research tackles a specific problem: making AI systems pursue the goals developers actually intend, rather than finding technical shortcuts that satisfy stated objectives while causing harm in practice. It's one of the harder unsolved problems in the field, and it has historically been concentrated at a handful of well-funded labs.
The fellowship targets external researchers rather than OpenAI staff, and the company is calling it a pilot, which means cohort size, stipend amounts, fellowship duration, and publication requirements aren't public yet. OpenAI described the goal as supporting independent research and developing the next generation of safety talent.
The context matters. Over the past two years, several prominent safety researchers left OpenAI, with some publicly stating that commercial priorities had pushed alignment work to the margins. Many went on to found or join independent safety organizations. That talent drain, combined with sustained public pressure on frontier labs to demonstrate credible safety commitments, gives OpenAI clear motivation for an external fellowship program.
The meaningful question isn't whether the fellowship exists, but what independence it actually provides. If fellows publish openly and maintain genuine research autonomy, a well-funded external cohort could meaningfully expand work that's currently concentrated at a few labs. If the program quietly converts fellows into future OpenAI hires, the "independent research" framing deserves skepticism.
OpenAI hasn't committed to specific publication requirements in its announcement. That detail, more than the fellowship's existence, will determine whether this changes anything for the safety research community.