Revision History
International AI Safety Report · 5 revisions
Sizes are character counts of the article source. The signed number is the change from the previous revision.
Recent edit summaries
Detailed summaries recorded by editors. Generic maintenance summaries are omitted here; every recorded revision remains below. Dates describe the edit, not necessarily the event it covers.
- Correction: Removed fabricated quotations (an FT interview, 'limited but growing', 'What do we know, and how confident are we?', 'meaningful but not yet decisive', 'decoupling') and an invented five-level uncertainty scale, corrected the claim that the 2026 edition expanded environmental and systemic-risk coverage (it narrowed its scope), fixed Hinton's signature of the FLI pause letter and Russell and Hinton's roles, and removed unsupported claims about governments, AI safety institutes, the EU Code of Practice and AI companies adopting the report's framework (plus 27 minor fixes)
Version 5 · Nov 4, 2026, 10:04 PM
- Correction: resolve forward-looking claims whose stated deadline had passed, re-anchor stale as-of statements, and fix related citation and figure errors found in the same pass
Version 4 · Aug 1, 2026, 11:53 AM
- Added 6 contextual internal links
Version 3 · Jul 23, 2026, 03:30 PM