Have OpenAI's updated frontier-training containment controls passed independent review?
External validation pending · initial record, no post-publication revision.
CORTIST / EVIDENCE INTELLIGENCE
What Changed?
Track genuine evidence revisions without erasing the original AI-claim record.
01 / Versioned evidence2026-10-23 / CL-0008 2026-10-21 / CL-0005 2026-10-19 / CL-0007 2026-10-16 / CL-0006 2026-10-15 / CL-0010 2026-10-14 / CL-0009
No post-baseline revisions yet.
The public Ledger began on 25 October 2026. Its first sources describe earlier developments, not subsequent change events. This view will record dated v2+ snapshots only when new material evidence actually changes an assessment.
0 verified status-changing revisions · 20 public cases (10 new initial baselines) · first snapshot 25 Oct 2026
Ten new case baselines · 5 November 2026
Cases CL-0011–CL-0020 are first-record snapshots, not status-changing updates to the original ten cases. No independent resolution was inferred. Inspect baseline additions and SHA-256 digests ↗
Recent source baselines
The entries below were documented at initial registration and must not be mistaken for newly completed promises.
What does the 3.7% decline in AI-exposed young-worker jobs prove?
Payroll pattern documented · initial record, no post-publication revision.
Will NVIDIA Vera's agent-sandbox throughput advantage hold for customers?
Rollout and replication pending · initial record, no post-publication revision.
Do Anthropic's junior hiring changes prove AI agents displaced workers?
Company workforce research · initial record, no post-publication revision.
Will METR publish independent safety evaluations after gaining access to Anthropic?
Independent METR report pending · initial record, no post-publication revision.
Do 134 IEA data-centre grid policies show that the measures work?
Policy inventory documented · initial record, no post-publication revision.