Boards build the oversight. Substance leads signal.
91% of Nasdaq-100 boards carry an operative oversight mechanism — a charter clause that says the committee shall oversee AI. Only 68% stop at an aspirational statement. Enacted runs 23pp ahead of stated.
Everywhere we look, mechanism leads messaging.
Each row is a cohort. The distance between the two dots is the substance premium: how far enacted oversight runs ahead of stated commitment.
Cohort by cohort
Gap = share with signal minus share with substance. Negative wherever coverage is real: more firms show an enacted board mechanism than an aspirational one.
| CohortCohort | FirmsFirms w/ board ev. | Gap | |||
|---|---|---|---|---|---|
| Nasdaq-100 | 66 | 91% | 68% | 6 | -23pp |
| Insurance | 39 | 90% | 51% | 4 | -38pp |
| Top-50 Tech | 49 | 92% | 61% | 4 | -31pp |
| Nasdaq Semis | 15 | 87% | 53% | 2 | -33pp |
| Top-50 Banks | 13 | 92% | 77% | 1 | -15pp |
| Nasdaq Software | 12 | 83% | 67% | 2 | -17pp |
| Internet & Media | 9 | 89% | 89% | 1 | 0pp |
| Healthcare | 7 | 100% | 71% | 0 | -29pp |
| Managed Care | 6 | 100% | 83% | 0 | -17pp |
Provenance
- Source table
- aigma_emissions (live Supabase)
- Filter
- channel = 'B' and superseded_by is null (current emissions only)
- Rollup unit
- firm. Has substance if it carries ≥1 SUB emission; has signal if ≥1 SIG.
- Script
- scripts/panel-study/compute-board-sub-sig.ts
All figures read the Board (DEF 14A) channel only, which over-represents substance because a proxy-statement charter is operative by construction. Read it as board-level oversight, not a company's overall AI maturity. Absence of a coded emission is absence of disclosure, not proof a mechanism does not exist.