METR

501(c)(3) · Berkeley, CA

Builds and runs evaluations that measure whether frontier AI systems can carry out long, autonomous tasks.

Epoch

501(c)(3)

Researches the trajectory of AI: compute, training data, benchmarks and economic impact.

Redwood

501(c)(3) · Berkeley, CA

Researches how to stay safe from AI systems that might act against their developers, known as AI control.

MATS

501(c)(3) · Berkeley, CA

A research and training program pairing emerging AI safety researchers with mentors.

FAR AI

501(c)(3) · Berkeley, CA

Does technical AI safety research, and runs workshops and a coworking space for the field.

Kairos

Runs programs that bring people into AI safety: SPAR research mentorship, and Pathfinder for university groups.

GovAI

501(c)(3) · London

Researches how to govern advanced AI, advises governments and labs, and trains people for AI governance careers.

MIRI

501(c)(3) · Berkeley, CA

Warns policymakers and the public about the danger of superhuman AI, and researches how to govern it.

Longview

London

Advises major donors and runs funds for reducing risks from AI, nuclear weapons and pandemics.

IAPS

501(c)(3)

A think tank researching policy for advanced AI, with a fellowship for people entering the field.

Lightcone

501(c)(3) · Berkeley, CA

We build beautiful infrastructure for truth-seeking and world-saving

BlueDot

Runs courses on AI safety, AI governance and biosecurity, and helps graduates into work on them.

Constellation

501(c)(3) · Berkeley, CA

Runs a research center in Berkeley for people working on AI safety, with fellowships and an incubator.

Center for AI Safety

501(c)(3) · San Francisco, CA

Research, field-building and advocacy to reduce societal-scale risks from AI.

Transluce

501(c)(3)

Ensure that AI oversight scales with AI capabilities

Sage

501(c)(3)

Builds interactive explainers and demos of AI (AI Digest, the AI Village) and forecasting tools like Fatebook.

1Day Sooner

501(c)(3)

Advocates for challenge-trial volunteers and for faster development of vaccines and treatments.

Palisade

501(c)(3)

Studies what the latest AI systems can do and how they behave, and shows the results to policymakers and the public.

Alignment Research Center

501(c)(3) · Berkeley, CA

Theoretical research on aligning future machine learning systems with human interests.

AI Futures Project

501(c)(3) · Berkeley, CA

A small research group forecasting the future of AI; the authors of AI 2027.

Forethought

Non-US

Navigate the transition to a world with superintelligent AI systems

Mox

Fiscally sponsored · San Francisco, CA

A coworking & events space in SF for AI safety, AI labs, EA charities & startups