Organizations shaping AI

Machine Intelligence Research Institute

Berkeley-based AI safety research organization focused on theoretical alignment, decision theory, and risks from advanced machine intelligence.

Redwood Research

AI alignment research organization focused on empirical safety work, interpretability, adversarial robustness, and failure modes in advanced models.

Alignment Research Center

Alignment Research Center develops methods for keeping powerful AI systems helpful and honest as their capabilities grow. Its work helped define scalable alignment and eliciting latent knowledge—ways to test whether models know more than they reveal—and its former evaluations team became METR.

Safe Superintelligence Inc.

AI safety-focused research company founded by Ilya Sutskever, Daniel Gross, and Daniel Levy. Focused exclusively on building safe superintelligence, with no products or near-term commercial pressure.

Apollo Research

Apollo Research tests whether advanced models can deceive evaluators, pursue hidden goals or evade human control. Its behavioral evaluations and interpretability research provide frontier labs and policymakers with concrete evidence about forms of model behavior that ordinary benchmarks miss.

Collective Intelligence Project

Research and governance incubator founded in 2023 by Divya Siddarth, Saffron Huang, and Jasmine Wang. Develops democratic processes — such as "alignment assemblies" — for steering transformative technologies like AI toward the collective good.

Conjecture

London-based AI alignment startup founded by Connor Leahy, Gabriel Alfour, and Sid Black. Spun out of the open-source collective EleutherAI to scale applied AI safety and alignment research.

Resolution

Resolution researches how to understand and control advanced AI systems before their reasoning becomes too complex for people to follow. Founded by Geoffrey Irving and Daniel Murfet, it combines alignment research with mathematics and interpretability aimed at making model behavior more legible.

FAR.AI

AI safety research nonprofit founded by Adam Gleave and Karl Berzins, based in Berkeley. Works on adversarial robustness, model evaluation, interpretability, and alignment, and convenes the AI safety research community.

Institute for Responsible Superintelligence

Develops formal, safe-by-design foundations for superintelligence, including precise safety objectives, mechanisms with analyzable guarantees, and proofs of concept that frontier AI developers can evaluate and adopt.

MATS Research

MATS runs fellowships that pair emerging researchers with experienced mentors and provide funding, research management and community for work on AI alignment, transparency and security. Its alumni have produced research and founded or joined organizations across the AI safety ecosystem.

Meaning Alignment Institute

Research institute founded by Joe Edelman, Oliver Klingefjord, and Ellie Hain studying how to align AI — as well as markets and democracies — with what people genuinely value, beyond engagement metrics.

Softmax

San Francisco AI alignment startup founded in 2025 by Emmett Shear (co-founder of Twitch and briefly interim CEO of OpenAI), Adam Goldstein, and David Bloomin. Researches "organic alignment," modeling how agents cooperate while keeping distinct identities.

An AI-safety research organization that applied singular learning theory—mathematical tools for understanding complex learning systems—to model interpretability and alignment. Its work and team have merged into Resolution.

Workshop Labs

Public-benefit AI research company developing user-aligned models, accessible post-training, privacy-preserving infrastructure, and systems designed for human collaboration.

AI Objectives Institute

AI Objectives Institute studies how the incentives inside AI systems, markets and bureaucracies can diverge from human interests. Its work combines alignment research with democracy, economic design, cooperation and structural-risk analysis.

Apart Research

Apart Research is a non-profit AI safety research organization that runs collaborative hackathons and fellowships to accelerate interpretability and alignment research.