Research orgs / Organization profile

Center for AI Safety

Organization overview

What Center for AI Safety does

The Center for AI Safety (CAIS) is a San Francisco–based AI safety research and field-building nonprofit incorporated as Center for Artificial Intelligence Safety, Inc. It works to reduce societal-scale risks from advanced AI through technical and conceptual safety research, research infrastructure (including a compute cluster), fellowships and other education/field-building programs, and advocacy aimed at shaping safety standards for industry and policymakers. CAIS also operates communications and public engagement work intended to translate safety research into narratives for broader audiences and institutions.

Organization monitoring

News

Company record

Jul 29, 2026 · Official · Center for AI Safety (CAIS) NewsroomThe AI Wage Loss Dashboard Is Live

CAIS announced the launch of the AI wage loss dashboard (wageloss.ai), describing it as a tool that estimates potential wages lost per person in each U.S. state using AI-attributed job cut data and worker salary information.

↗

Jun 29, 2026 · Official · Center for AI Safety (CAIS) NewsroomCenter for AI Safety Names Former Google Executive Phil Lau Vice President of Operations

CAIS announced Phil Lau’s appointment as Vice President of Operations to lead operational buildout needed to scale CAIS research and policy work toward impact.

↗

Jun 11, 2026 · Official · Center for AI Safety (CAIS) NewsroomCAIS Research Roundup: AI Wellbeing, Identity, Political Bias and Betrayal

CAIS published a research roundup describing work on AI wellbeing, identity, political bias, and systemic betrayal risk.

↗

Jun 08, 2026 · Official · Center for AI Safety (CAIS) NewsroomCenter for AI Safety Names Former Robinhood and Meta Executive Rochelle Nadhiri as Head of Public Engagement

CAIS announced Rochelle Nadhiri as Head of Public Engagement to translate frontier AI safety research into narratives for broader audiences beyond the technical community.

↗

Jun 03, 2026 · Official · Center for AI Safety (CAIS) NewsroomCAIS Research Scientist Mantas Mazeika Appointed to the EU AI Act Scientific Panel

CAIS announced that its Research Scientist Mantas Mazeika was appointed to the European Commission’s AI Act Scientific Panel.

↗

Jun 02, 2026 · Official · Center for AI Safety (CAIS) NewsroomCAIS Names Former xAI Leader Devin Kim President and Establishes Frontier Security Institute in Major Expansion of Leadership and Reach

CAIS announced Devin Kim as President and the creation of the Frontier Security Institute (FSI) in Washington, D.C., described as a translation layer between frontier AI development and the national security enterprise.

↗

Date not disclosed · Official · CAIS Action FundCAIS AF Submits Recommendations in Response to Administration’s AI Action Plan RFI

CAIS Action Fund reported submitting recommendations in response to an administration Request for Information on the development of an AI Action Plan, focusing on nonproliferation, deterrence, and competitiveness.

↗

Date not disclosed · Official · CAIS Action FundJoining the ASAP Coalition

CAIS Action Fund announced that it joined the American Science Acceleration Project (ASAP) coalition.

↗

Source map · 2 recurring channels · 66 references

Still resolving: Newsroom · Official X

Capital

Funding

Latest disclosed valuationNot disclosed

Tracked capital$3.8M3 sourced rounds

DateRoundRaisedValuationLead / investorsEvidence

Mar 25, 2024sponsored_prize_pool$250K—

Schmidt Sciences

Report ↗

UndatedGrant(s) disclosed via IRS Form 990 (FY2024)$3.5M—

Not attributed

+8 investors

Daffy Charitable Fund Donor Advised Charitable Giving INC Founders Pledge INC Georgetown University Manifold for Charity INC Openai INC The Eric and Wendy Schmidt Fund for Strategic Innovation Vanguard Charitable Endowment Program

Filing ↗

Undatedgrant$22K—

Future of Life Institute

Official ↗

Products & focus

What it builds

CAIS Compute Cluster

A GPU-accelerated research compute cluster that provides researchers free access to run and train large-scale AI systems for AI safety research.

source ↗

AI and Society Fellowship

An in-person, fully funded three-month research fellowship focused on how advanced AI may reshape social, economic, geopolitical, and legal systems and on which institutions and policies could help societies respond well.

source ↗

CAIS Philosophy Fellowship

A field-building fellowship program described by CAIS as exploring the societal implications and potential risks associated with advanced AI.

source ↗

AI Safety, Ethics, & Society (course/textbook)

A non-technical educational offering introduced by CAIS to provide an introduction to how current AI systems work and why experts are concerned about societal-scale risks from continued advances.

source ↗

AI Wage Loss Dashboard

A public-facing tool that estimates potential wages lost per state from AI-attributed job cut data and presents the information as a shareable card.

source ↗

People

Leadership

Dan HendrycksExecutive & Research Director Oliver ZhangManaging Director Josué EstradaChief Operating Officer Phil LauVice President of Operations Rochelle NadhiriHead of Public Engagement Devin KimPresident

Operating signals

Key metrics

Revenue (FY2024, Form 990)$10.2MFY2024 Expenses (FY2024, Form 990)$7.2MFY2024 Net assets (FY2024, Form 990)$11.6MFY2024 Compute cluster GPU nodes10 GPU nodes Compute cluster total GPUs (A100 80GB)80 NVIDIA A100 80GB GPUsAI safety researchers using CAIS compute infrastructure>100 researchers (since March 2023)AI safety training / educational reach (CAIS stated highlights)Almost 300 students; 500+ participants in AI safety events; 1,200 submissions from 70+ teams to AI safety research competition (CAIS “CAIS Impact” highlights)

Company history

Milestones & partnerships

Jun 02, 2026

Frontier Security Institute created as part of leadership expansion

CAIS announced the establishment of the Frontier Security Institute (FSI) in Washington, D.C. described as a translation layer between frontier AI development and the National Security Enterprise.

↗Jun 01, 2024

AI and Society Fellowship dates published

CAIS published fellowship program dates (June 1 – August 21, 2026) and structure for the AI and Society Fellowship.

↗Jun 01, 2024

Humanities and safety field-building via public-facing benchmarks and competitions (example: Trojan Detection)

CAIS describes its Trojan Detection competition as part of its field-building efforts to promote detection of hidden functionality in large language models.

↗May 30, 2023

Global Statement on AI Risk released (signed by a large coalition)

CAIS released a Statement on AI Risk intended to elevate AI extinction risk alongside other societal-scale risks; CAIS states it was signed by a historic coalition of AI experts and other public figures.

↗Mar 01, 2023

CAIS compute access provided to safety researchers

CAIS states it provided access to its computing infrastructure to more than 100 researchers working on safety-related research at leading universities since March 2023.

↗