AGI Organisations to Follow
The labs, institutes, and policy bodies worth tracking.
Frontier lab · USA
OpenAI
Builder of the GPT family and ChatGPT; mission-stated focus on AGI for the benefit of humanity.
Visit site ↗Frontier lab · UK / USA
Google DeepMind
Google's combined frontier AI lab, builder of Gemini and AlphaFold.
Visit site ↗Frontier lab · USA
Anthropic
Frontier lab focused on AI safety research; builder of the Claude family.
Visit site ↗Frontier lab · USA
Meta AI (FAIR)
Meta's research arm, leading on the open-weight Llama family.
Visit site ↗Frontier lab · USA
xAI
Frontier lab building the Grok model family.
Visit site ↗Frontier lab · France / EU
Mistral AI
European frontier lab focused on open-weight and enterprise models.
Visit site ↗Frontier lab · USA
Microsoft AI
Hosts and integrates frontier models across Azure, Copilot, and Windows.
Visit site ↗Frontier lab · USA
NVIDIA Research
Compute, simulation, and model research powering most frontier training runs.
Visit site ↗Safety / research · USA
MIRI (Machine Intelligence Research Institute)
Long-standing research nonprofit focused on technical alignment and existential risk.
Visit site ↗Safety / research · USA
Alignment Research Center (ARC)
Theoretical and applied alignment research, including dangerous-capability evaluations.
Visit site ↗Safety / research · UK
Apollo Research
Evaluations of deceptive behaviour and dangerous capabilities in frontier models.
Visit site ↗Safety / research · USA
Center for AI Safety (CAIS)
Research, advocacy, and field-building focused on catastrophic AI risk.
Visit site ↗Policy / governance · UK
UK AI Safety Institute
Government institute conducting frontier-model evaluations on behalf of the UK.
Visit site ↗Policy / governance · USA
US AI Safety Institute (NIST)
NIST-hosted institute coordinating US frontier-model safety work.
Visit site ↗Policy / governance · EU
EU AI Office
Implements and enforces the EU AI Act and coordinates with national regulators.
Visit site ↗Policy / governance · International
OECD AI Policy Observatory
Multilateral hub for AI policy data, the OECD AI Principles, and country comparisons.
Visit site ↗Academic · USA
Stanford HAI (Human-Centered AI Institute)
Stanford's interdisciplinary AI institute and publisher of the AI Index Report.
Visit site ↗Academic · USA
MIT CSAIL
MIT's flagship AI and computer-science lab.
Visit site ↗Academic · USA
Berkeley CHAI (Center for Human-Compatible AI)
Stuart Russell's lab on provably beneficial AI.
Visit site ↗Civil society · International
Partnership on AI
Multistakeholder organisation developing AI best practices across industry, civil society, and academia.
Visit site ↗Civil society · USA
AI Now Institute
Research institute focused on the social implications and concentrations of power in AI.
Visit site ↗