Reports
Findings and the analyses they headline, built on the panel study. Each finding shows its number honestly (live, provisional, or pending) and links to where the analysis lives.
Among Nasdaq-100 firms with board AI disclosures, most show an operative oversight mechanism.
Of the "AI-governance" roles tagged so far, most are build roles. Only a minority actually govern.
For every AI-governance job posted in 2025, employers posted 103 asking for AI skills.
Of career sites hiring for AI for five years or more, 69% have never posted an AI-governance job title.
A firm that claims L3 practice shows it within two years just over half the time.
Insurers leave 60% of their AI-governance claims undemonstrated, against 33% for the Nasdaq-100.
Our media coder can tell what a firm did from what it said. Early check, one firm.
Of 35 consulting-firm cells we scored as dark, 20 had governance documents we had never collected, mostly PDFs.
The analyses that findings roll up into. A report links out once it ships a route.
On the board (DEF 14A) channel, how many firms carry an operative oversight mechanism versus an aspirational statement, cohort by cohort. The direction is not the same in every cohort. Reads live from current B emissions.
Board channel only, which over-represents substance because a proxy-statement charter is operative by construction.
How firms hire to govern AI across the covered panel: the govern-vs-build split among captured AI roles, the roles that dominate, and a baseline for tracking the mix forward. Industry cuts appear once each has enough firms.
Mixed bank and tech panel, not industry-weighted. LLM-coded ARISA role family; not yet validated against human labels.
Over a historical vendor job-posting corpus: how AI hiring differs between US- and EU-located requisitions, plus an exploratory look at governance-adjacent hiring by EU- vs US-headquartered firm pairs against the EU AI Act calendar (no causal inference). Reads live from the deterministic tag surface.
Descriptive, scoring-invisible vendor corpus; read shape not level (capped exports, left-censoring). Governance-adjacent by half-year is a deliberate, labeled exploratory relaxation — the frozen DiD outcome (factor_target) stays withheld pending the ai-act-hiring-did pre-registration.
- Cross-Cohort Governance MovementPlannedIndex
Quarter-over-quarter shifts in the governance index across cohorts, with the evidence behind each move.
- Model-Risk Apparatus vs. Hiring SignalPlannedIndex
Whether banks that build a deeper model-risk apparatus also hire more governance roles, and whether either tracks the index.
- Governance Index Quarterly LetterPlannedBriefing
An editorial recap of where the cohort stands, what moved this quarter, and what to watch next.