Skip to content
Longterm Wiki
All Source Checks
Division

Safeguarded AI Programme

confirmed95% confidence

1 evidence check

Last checked: 3/26/2026

The source text directly confirms all three key fields: (1) name 'Safeguarded AI Programme' is not explicitly stated in the excerpt but is strongly implied by the context of 'this programme' discussing AI safety with quantitative guarantees; (2) type as 'program-area' is confirmed by the description of the programme being 'split into three technical areas (TAs)'; (3) status as 'active' is confirmed by multiple references to current projects, ongoing decisions, and active management ('Current TA3 Phase 1 Projects – All existing TA3 Phase 1 projects will continue'). The only minor uncertainty is that the exact name 'Safeguarded AI Programme' is not explicitly stated in this excerpt, though the context strongly supports it.

Evidence — 1 source, 1 check

confirmed95%Haiku 4.5 · 3/26/2026
Found: The source text describes a programme with the goal to 'usher in a new era for AI safety' and discusses its structure across three technical areas (TA1, TA2, TA3). The programme is actively managing p

Note: The source text directly confirms all three key fields: (1) name 'Safeguarded AI Programme' is not explicitly stated in the excerpt but is strongly implied by the context of 'this programme' discussing AI safety with quantitative guarantees; (2) type as 'program-area' is confirmed by the description of the programme being 'split into three technical areas (TAs)'; (3) status as 'active' is confirmed by multiple references to current projects, ongoing decisions, and active management ('Current TA3 Phase 1 Projects – All existing TA3 Phase 1 projects will continue'). The only minor uncertainty is that the exact name 'Safeguarded AI Programme' is not explicitly stated in this excerpt, though the context strongly supports it.

Debug info

Record type: division

Record ID: CQswIy5lQS