Anthropic opens a paid safety research residency for professors, founders and senior engineers without ML backgrounds
Placements last six weeks, three months or six months with an Anthropic mentor and a compute budget; applications close on November 22, 2026, and the number of places is not stated.
Why it matters
Anthropic is widening its safety research intake beyond machine learning specialists, arguing that framing a research agenda and judging what is worth doing now matter more than ML training. The residency builds on the Anthropic Fellows Program, whose participants produced work on subliminal learning, the Petri auditing tool and persona vectors. Residents work on problems set by Anthropic researchers, preferably from San Francisco, with remote participation possible. Watch for how many senior people from other fields take up the offer before the November 22 deadline.
Key facts
- Applications close November 22, 2026
- Placements of six weeks, three months or six months
- Paid, with an Anthropic mentor and a dedicated compute budget
- No machine learning background required
- Open to professors, technical founders and staff-level engineers
- Applications for the US and the UK