Observed arrival · 2026-10-24
Lab Canary’s AI-lab speak-up scorecard
A public-evidence scorecard examines how AI labs’ employment terms and speak-up policies address criticism and whistleblowing.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- People assessing speak-up protections at AI labs
- Worth noticing
- A public statement earns at most “Partly”; “Meets” requires written policy or contract wording.
Field notes
The scoring rules distinguish evidence quality: “Meets” requires written policy or contract language, while a public statement alone earns no more than “Partly.” The maker says he has applied to Anthropic, OpenAI, and Google DeepMind fellowships, receives no payment for the site, and invites corrections; “Not public” is explicitly a score for openness, not a finding that terms are bad.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
◉HumanPersonal, local, civic, or handmade
One card from the complete issue