FAR.AI

About FAR.AI

FAR.AI is a non-profit research institute focused on AI safety and alignment. It conducts technical research, red-teaming, and evaluations of advanced AI, and organizes events and programs to build the global field of trustworthy AI.

FAR.AI is a non-profit research institute working to ensure advanced AI is safe and beneficial. Founded in July 2022, it conducts technical research on AI alignment topics such as adversarial robustness, mechanistic interpretability, and mitigating AI deception. The organization also performs red-teaming and pre-deployment evaluations of frontier models for developers like OpenAI and for governments such as the EU AI Office.

FAR.AI builds the field through events and programs. It hosts the Alignment Workshop series and specialized workshops, and it runs FAR.Labs, a co-working space in Berkeley for AI safety researchers, along with a grantmaking program that supports academic and independent researchers working on critical AI safety problems.

FAR.AI is supported by donations and grants from supporters, and it provides evaluation and testing services to frontier AI developers and government bodies.