AI Lab Watch

I stopped maintaining this website in September 2025. Work on what AI companies should do in terms of safety and what they are doing is now being done by METR, Guidelight, Midas, and others.

			*
Weighted score	28%	20%	18%	5%	4%	3%	1%
Risk assessment	44%	29%	34%	6%	6%	1%	0%	27% weight
Scheming risk prevention	4%	8%	3%	2%	2%	2%	2%	21% weight
Boosting safety research	70%	52%	35%	25%	0%	13%	8%	14% weight
Misuse prevention	14%	4%	9%	0%	1%	0%	0%	12% weight
Prep for extreme security	3%	5%	0%	0%	0%	0%	0%	12% weight
Risk info sharing	35%	13%	32%	0%	28%	0%	0%	8% weight
Planning	14%	26%	0%	0%	0%	1%	0%	6% weight

Up to date as of September 2025

Overall score

Anthropic

DeepMind

OpenAI

Anthropic

Prep for extreme security 3%

Plan for SL5

Red-team resilience

10%

Practices: isolated network

10%

Practices: secure developers' machines

25%

Practices: multiparty controls

10%

Practices: secure boot

Track record

AI companies should prepare to protect model weights and code by the time AI massively boosts R&D, even from top-priority operations by the top cyber-capable institutions.

Compare all companies on prep for extreme security

Plan for SL5

Click to show details/rubric

Red-team resilience

Practices: isolated network

10%

Click to show details/rubric

Practices: secure developers' machines

10%

Click to show details/rubric

Practices: multiparty controls

25%

Click to show details/rubric

Practices: secure boot

10%

Click to show details/rubric

AI Lab Watch

Categories

Companies

Resources

Blog

About

Anthropic

Prep for extreme security 3%

Plan for SL5

Red-team resilience

Practices: isolated network

Practices: secure developers' machines

Practices: multiparty controls

Practices: secure boot

Track record