A residency for AI safety researchers to collaborate and work alongside the AI safety community in Beijing.
Spend three months at a new AI safety hub in Beijing working with the AI safety research community in China on frontier AI safety. The hub is led by Chief Scientist Professor Andrew Yao (China’s only Turing Award winner) and co-directed by Professor and Vice Dean Xu Wei and Dr. Sören Mindermann, working with leading Chinese AI safety experts and the safety teams of Chinese frontier labs.
We are recruiting a small, inaugural cohort of around five to ten researchers — mostly technical AI safety researchers. The programme is highly selective and high-touch — we will work closely with each resident to make their residency productive and impactful.
This new AI safety hub in Beijing is a daily working community for safety researchers from Chinese universities and frontier labs. It is led by Chief Scientist Professor Andrew Yao, China’s only Turing Award winner and a key voice on frontier AI risks. It is co-directed by Professor Xu Wei (IISS/CollegeAI) and Dr. Sören Mindermann (Cambridge/International AI Safety Report), working with a range of Chinese AI safety experts and safety teams of AI companies.
The Residency is the hub’s channel for international AI safety researchers to spend time living and collaborating with the AI safety research community in China — working on their own research with full access to the hub community, pursuing joint projects, and contributing to a sustained channel for international cooperation on safety research.
This is the inaugural cohort. We are deliberately keeping it small and selective. Residents will have real input into how the programme takes shape, and should be comfortable working with us as some details continue to settle.
The Residency’s operations are supported and facilitated by the Safe AI Forum (SAIF), a US-based non-profit that works on international coordination on AI safety — through the International Dialogues on AI Safety, research collaborations, and other programmes since 2023.
We expect most of the cohort to be experienced technical AI safety researchers with a publication record, here to collaborate and potentially mentor. Technical researchers work on alignment, interpretability, evaluations, scalable oversight, control, dangerous capabilities, or related problems — publications or comparable outputs (including open-source projects and technical reports) expected.
We are also inviting a small number of governance applicants — including work on technical governance (verification, compute governance). Future cohorts may open further to more governance candidates. Non-traditional backgrounds are welcome — we do not require a PhD.
The initial application is intentionally short — we’d rather hear the rest of the story in conversation.
Have you been referred? If a member of the team has invited you to apply or referred you to the fast track, use the short fast-track form instead.
If you can’t join us for a full residency but would still be interested in visiting the hub, let us know. Fill out our short interest form and we’ll be in touch in late July with updates about shorter-term visits and other programming.
Register your interest