The Labs
The organisations behind AI — the safety labs and nonprofits trying to make it go well, and the open-weight labs shipping the models. RSS · 64 episodes
The Centre for Long-Term Resilience — Extreme Risk as a Civil-Service Brief
Sep 17, 2026
Simplex — The Physics of What a Transformer Believes
Sep 16, 2026
AE Studio — The Consultancy That Pays For Its Own Alignment Lab
Sep 15, 2026
Gradual Disempowerment — The Catastrophe In Which Nothing Rebels
Sep 14, 2026
Lightcone & Lighthaven — The Rationalist Commons
Sep 13, 2026
The Frontier Model Forum — When the Labs Write the Standard
Sep 12, 2026
The UN's Scientific Panel on AI — Forty Experts and No Power
Sep 11, 2026
Concordia AI — The Safety Org Inside China
Sep 10, 2026
The Center on Long-Term Risk — The Worst Case Nobody Else Works On
Sep 09, 2026
Atlas Computing — The Case for Rules Instead of Values
Sep 08, 2026
SecureBio — Measuring How Much a Model Helps a Virologist
Sep 07, 2026
SaferAI — Grading the Graders
Sep 06, 2026
Haize Labs and Virtue AI — The Second Wave of Safety Startups
Sep 05, 2026
Japan's AI Safety Institute — The Documentation Shop Asked to Become an Eval Lab
Sep 04, 2026
Truthful AI — The Lab That Found Misalignment Generalises
Sep 03, 2026
Open Philanthropy, Now Coefficient Giving — The Foundation Behind the Field
Sep 02, 2026
The Simon Institute — Lobbying the United Nations Instead of Congress
Sep 01, 2026
The Center for AI Policy — The Advocacy Shop That Ran Out of Money
Aug 31, 2026
The Cooperative AI Foundation — The Risk That Needs Two Agents
Aug 30, 2026
OpenAI's Safety Teams — Three Disbandings and a Counter
Aug 29, 2026
Anthropic Alignment Science — Publishing the Bad Numbers
Aug 28, 2026
Google DeepMind AGI Safety and Alignment — Doing It Inside the Building
Aug 27, 2026
Irregular — The Frontier AI Security Lab That Sells to the Labs
Aug 26, 2026
Andon Labs — Testing Whether a Model Can Run a Business Without Falling Apart
Aug 25, 2026
Eleos AI Research — Taking the Question of AI Suffering Seriously
Aug 24, 2026
Americans for Responsible Innovation — The Safety Lobby Learns to Campaign
Aug 23, 2026
Carnegie's China AI Initiative — The Other Half of the Board
Aug 22, 2026
The Institute for Law & AI — Writing the Legal Plumbing
Aug 21, 2026
Apart Research — Turning Weekends Into Researchers
Aug 20, 2026
Gray Swan — The Red Team Every Frontier Lab Buys
Aug 19, 2026
Conjecture — The Alignment Startup That Called Time On Itself
Aug 18, 2026
CAISI — What America Kept When It Dropped the Word "Safety"
Aug 17, 2026
Safeguarded AI — Britain Bets Fifty-Nine Million on Mathematical Proof
Aug 14, 2026
Palisade Research — The Demo Is the Argument
Aug 13, 2026
CeSIA — Building an AI Safety Scene in France
Aug 12, 2026
CHAI — Stuart Russell's Bet That the Objective Is the Bug
Aug 11, 2026
The Future Society — Lobbying the AI Act Into Existence
Aug 10, 2026
The EU AI Office — The Regulator That Has to Build Its Own Eval Lab
Aug 09, 2026
Ought and Elicit — Betting on Process Over Outcome
Aug 08, 2026
EleutherAI — The Open-Source Lab That Argues Openness Is Safety
Aug 07, 2026
The Ada Lovelace Institute — Britain's Public-Interest Referee
Aug 06, 2026
AI Now — The Critique From the Other Side
Aug 05, 2026
RAND — The Establishment Takes the Brief
Aug 04, 2026
IAPS — Policy for the Transition to Advanced AI
Aug 03, 2026
Resolution — Leaving the State to Automate Alignment
Aug 02, 2026
Timaeus — Betting Alignment on Pure Mathematics
Aug 01, 2026
Goodfire — Interpretability as a Business
Jul 31, 2026
LawZero — Bengio Builds the Non-Agentic Alternative
Jul 30, 2026
The UK AI Security Institute — The State Builds an Eval Lab
Jul 29, 2026
Epoch AI — Counting the Trend Lines
Jul 28, 2026
Transluce — Building the Public Instrument
Jul 27, 2026
FAR.AI — Capability Is Not Robustness
Jul 26, 2026
CSET — The Hundred-Million-Dollar Bet on Washington
Jul 25, 2026
The Center for AI Safety — From the Extinction Statement to the Pentagon
Jul 24, 2026
GovAI — The Think Tank That Wrote the Rules
Jul 23, 2026
The Alignment Research Center — Theory for the Hard Part
Jul 22, 2026
Apollo Research — Hunting for Scheming and Deception
Jul 19, 2026
Redwood Research — Betting on AI Control
Jul 17, 2026
METR — Measuring What Frontier Models Can Actually Do
Jul 16, 2026
Moonshot & Zhipu — Beijing's Open-Model Contenders
Jul 04, 2026
Qwen — Alibaba's Open-Weight Juggernaut
Jul 03, 2026
DeepSeek — The Hedge Fund That Open-Sourced a Frontier Model
Jul 02, 2026
Black Forest Labs — The Black Forest's Open Image Lab
Jul 01, 2026
Mistral AI — Europe's Open-Weight Champion
Jun 30, 2026
