Development

Every model since one release scores perfectly on a misalignment eval

development2026-05-09

Claim

Anthropic notes that since Claude Haiku 4.5 every Claude has scored perfectly on agentic misalignment, the same evaluation Opus 4 once failed 96% of the time.

Reported metric: 100% vs 96% failure

Reporting date: 2026-05-09 (not necessarily the event date)

Sources

Attribution and reporting linked by this record; links do not imply independent verification.

Welcome to May 9, 2026 (theinnermostloop.substack.com)

Knowledge graphopen in browser →

Reported in

Welcome to May 9, 2026

Actor

Anthropic

Evidence for

Supersedes

A safety institute counts 700 real-world cases of scheming

Referenced by

Welcome to May 9, 2026

Cite or correct this record

Citation formatCitation text

Every model since one release scores perfectly on a misalignment eval
Innermost Loop KB — secondary record
Reporting date: 2026-05-09
KB permalink: https://nicholsn.github.io/innermost-loop-kb/developments/2026-05-09-perfect-scores-on-the-eval-that-once-failed
Claim: Anthropic notes that since Claude Haiku 4.5 every Claude has scored perfectly on agentic misalignment, the same evaluation Opus 4 once failed 96% of the time.
Attributed sources:
- Welcome to May 9, 2026 — https://theinnermostloop.substack.com/p/welcome-to-may-9-2026
This citation does not imply independent verification of linked-source claims.

Suggest a correction — opens a GitHub draft for you to review and submit.