
Nonprofit · Berkeley, United States
FAR.AI
A nonprofit AI security research organization that builds safety solutions, tests frontier models and convenes the field.
Berkeley, United States
- Founded
- 2022
- Type
- Nonprofit
- Based in
- Berkeley, United States
- Key papers
- 3
- Featured projects
- 4
About
FAR.AI describes itself as an AI security research and adoption non-profit. It was co-founded by Adam Gleave, who is CEO, and Karl Berzins, who is president.
Its work spans in-house research agendas, applied research that validates ideas through near-frontier-scale prototypes, pre-deployment testing and consulting with frontier developers, government capacity building, field-building events and grantmaking, and public safety evaluations. Its GitHub organisation, AlignmentResearch, publishes code from this work.
Mission and ambitionsFAR.AI aims to close the gap between safety research and real-world adoption so that advanced AI is trustworthy, secure and beneficial.
Known for
- 01
AI safety and security research
- 02
Pre-deployment testing of frontier models
- 03
Convenings and field-building events
Open source
Tuned lens tools
Adversarial attacks on Go AIs
Vision-language models as rewards
Learned planner research
Research papers
Notable releases
Openness
MIT and Apache-2.0 research code on GitHub.
People
- Adam GleaveCo-founder and CEO
- Karl BerzinsCo-founder and President
- Kellin PelrineHead of Technology, Frontier Safety
- Boston NyerChief Operating Officer
- Chris CundyMember of Technical Staff
Listed from the lab's own pages and public profiles.
Contact and community
Via website.
Sources
Compiled from the lab's public pages, Hugging Face and GitHub. Last checked 2026-11-17. Something wrong or out of date? Tell us on Discord.



