Frontier Safety Ledger: weekly scorecard rates 10 AI labs across 7 safety categories

ItsAliChaudhry · x · 2026-11-16

Veracious AI (independent, unaffiliated) launched the Frontier Safety Ledger, a weekly-updated scorecard scoring 10 frontier AI developers out of 100 across 7 weighted categories: risk assessment, scheming prevention, safety research, misuse, security, information sharing, and planning. Every cell links to reasoning and dated sources; there's a filterable recent-shifts feed, an AI incidents tracker and timeline, and cross-references to the FLI AI Safety Index and SaferAI. Scores are editorial estimates from public information, with absence of evidence treated as absence of proof.

Original post →

More from Safety

Safety channel →