AI-REVIEWED RESEARCH · VERSION 1.1
Five dimensions, evidence gates
The weighted index produces a proposed 0–5 band. Gates can reduce a score and hold publication. Missing evidence stays grey. Source cutoff: 14 October 2026.
| Dimension | Weight |
|---|
| D1 · Legal incorporation | 30% |
| D2 · Scope and applicability | 15% |
| D3 · Transition maturity | 10% |
| D4 · Audited implementation | 30% |
| D5 · Supervisory response | 15% |
Complete dimensions are rounded to the nearest band at 0.5, 1.5, 2.5, 3.5 and 4.5. Confidence C is provisional and excludes score 5. Confidence D excludes a headline score.
What each evidence layer means
Domestic statutory statements, multinational group accounts and prudential returns describe different reporting populations. OHADA and CIMA rules are recorded by layer. A regional rule or a parent-company filing does not prove every local insurer applies IFRS 17.
Publication requirements
- Research must be Ready with confidence A, B or C.
- D1 ≤2 caps the headline at 2.
- D4 ≤1 caps the headline at 3.
- Score 5 requires D1–D4 ≥4 and D5 ≥3.
- Material unresolved conflicts cap the score at 3 and confidence at C.
- Scores 4–5 require a recorded second independent review.
Independent AI review and publication
Methodology 1.1 permits a fresh, independently sealed AI review to satisfy the second-review gate for this research publication. The five dimensions, weights and scoring criteria above are retained. A review can approve or hold a candidate; approval alone does not create a published score.
Fresh reviews cover nine candidates. Each country shows its review decision, date and coverage limitations. Earlier AI review summaries are identified separately and do not count as fresh reviews. Published confidence C scores remain provisional.
All research and reviews shown here were produced with AI. They do not provide human sign-off, external professional assurance or a regulatory determination. This dashboard is a dated research snapshot.
Read the public review summary →
