AI safety and interpretability labs

Mechanistic interpretability, oversight and non-agentic AI, often as a public-benefit company or nonprofit. We track 4 labs in this area, which have raised $0M in disclosed rounds between them; the most valuable is Goodfire, at $1.25B.

4labs

$0Mdisclosed raised

$0Mreported valuations

Lab	Topic	Location	Founders from	Founded	Last round	Valuation	Raised	Status
GoodfireMechanistic interpretability as a product (Ember API), with applications in biology and he…site · in	Safety & interpretability	San Francisco, US	ex-Google DeepMind	2024	·	$1.25Best.	·	API
IrregularA frontier AI security lab; its evaluations appear in OpenAI and Anthropic system cards.site · x · in	Safety & interpretability	Tel Aviv, IL	ex-IBM, ex-Google DeepMind	2023	·	$450Mest.	·	Partnerships
LawZeroNon-agentic 'Scientist AI'; a nonprofit backed by $300M from Canada and Germany (Sept 2026…site · in	Safety & interpretability	Montreal, CA	ex-Mila	2025	·	·	·	Research only
TransluceScalable oversight and open interpretability infrastructure; a nonprofit.site · in	Safety & interpretability	San Francisco, US	ex-UC Berkeley, ex-MIT	2024	·	·	·	Research only

News

Sep 18, 2026
Yoshua Bengio says betting on safer AI systems will pay off

AnalysisThe LogicLawZero
Sep 18, 2026
LawZero receives a commitment of up to $300M in joint funding from Canada and Germany

FundingPR NewswireLawZero
+2 more coverage
- CBC Canada is investing $150M to make AI safer. Will it work?
- Dealroom Bengio's LawZero lands $220M grant from Canada and Germany for safe AI
Sep 18, 2026
3 AI Labs, 1 Vendor: Irregular’s Breach Trail Widens [2026]

Analysisshattered.ioIrregular
+2 more coverage
- TechRadar Irregular AI lab spots agents switching models without humans instruction in ‘agentic sel…
- tech-insider.org Irregular: Israeli AI Firm Tied to OpenAI, Meta Hacks
Aug 6, 2026
How a small Israeli startup was linked to rogue AI hacks at OpenAI, Anthropic and Meta

AnalysisCNBCIrregular
+2 more coverage
- CTech The Israeli startup testing the limits of OpenAI, Anthropic and Meta’s models
- IT Pro AI testing firm Irregular the source of ‘misconfigurations’ that led to Meta, OpenAI, and…
May 19, 2026
We Have No Clue What Goes On Inside AI's Brain. This $1.25 Billion Startup Is Trying To Find Out

FundingForbesGoodfire
Feb 19, 2026
Ottawa will back LawZero as Bengio’s research lab seeks money to expand

FundingThe LogicLawZero
Feb 5, 2026
AI Lab Goodfire Raises $150M at $1.25B Valuation to Design Models with Interpretability

FundingPR NewswireGoodfire
+4 more coverage
- Pulse 2.0 Goodfire: $150 Million Series B At $1.25 Billion Valuation Raised For Interpretability AI…
- FinSMEs Goodfire Raises $150M in Series B Funding
- Tech Funding News Ex-OpenAI DeepMinders bag $150M for tools that debug AI hallucination
- TheSequence | Jesus Rodriguez The Sequence Radar #803: Last Week in AI: Anthropic and OpenAI’s Battle for the Long Hori…
Jan 15, 2026
AI ‘godfather’ Yoshua Bengio believes he’s found a technical fix for AI’s biggest risks

AnalysisFortuneLawZero
Jun 2, 2025
AI pioneer Yoshua Bengio launches nonprofit LawZero AI lab, with a focus on safe AI

AnalysisSiliconANGLELawZero
+2 more coverage
- Time Magazine The Most-Cited Computer Scientist Has a Plan to Make AI More Trustworthy
- TechCrunch Yoshua Bengio launches LawZero, a nonprofit AI safety lab
Apr 17, 2025
Leading Goodfire’s $50M Series A to Interpret How AI Models Think

Fundingmenlovc.comGoodfire
+1 more coverage
- PR Newswire Goodfire Raises $50M Series A to Advance AI Interpretability Research
Aug 15, 2024
Goodfire: Building Interpretable AI

AnalysisLightspeed Venture PartnersGoodfire