# Fortune 500 Chief AI Officer census

A company-by-company check of which Fortune 500 companies have a Chief AI Officer,
built from press and company sources rather than self-reported profile titles, and
set against the figure of 43% that circulates online.

Snapshot: 17 September 2026. Universe: the 2026 Fortune 500 list, 500 companies,
in `installmap.com/data/universes/fortune500/companies.csv`.

**The headline number is a floor.** 24 of 500 is the count of companies with press
or company evidence of a current, enterprise-wide Chief AI Officer on the snapshot
date. It is not a claim that the other 476 have nobody. See Limitations in
METHOD.md.

**Every company was read.** The cached search results for all 500 companies were
read by a person, with no proximity filter, and every row in `census.csv` carries a
`read_note` saying what its results contained.

## Headline

| Press status | Companies (of 500) | Share |
|---|---|---|
| appointed (enterprise-wide, current) | 24 | 4.8% |
| combined title (Chief Data and AI Officer and similar, enterprise-wide) | 30 | 6.0% |
| unverified (only a vendor page, directory, conference bio, award bio or LinkedIn claims it) | 39 | 7.8% |
| divisional (title names a unit, region or subsidiary) | 14 | 2.8% |
| former (person left or role folded) | 12 | 2.4% |
| none | 381 | 76.2% |

**24 of 500 Fortune 500 companies (4.8%)** have an enterprise-wide Chief AI Officer
on press or company evidence. The count is press evidence only: a title that exists
only on a LinkedIn profile is `unverified` wherever it appears, which is why
Fastenal, Paccar and WESCO International are not in it. Counting divisional holders and combined titles as well, the
widest number the evidence supports is **68 of 500 (13.6%)**.

The circulating figure is 43%.

## Ranks

Two ranks are shared, because Fortune's own list ties them on revenue and skips the
next number: Applied Materials and Hartford Insurance Group are both 160 and there
is no 161; Community Health Systems and Huntington Bancshares are both 350 and
there is no 351. The 500 companies carry 498 distinct ranks.

## Files

| File | Rows | What it is |
|---|---|---|
| `census.csv` | 500 | One row per Fortune 500 company: rank, name, domain, press_status, person, title, date, source URL, source type, LinkedIn tier, `query_pass`, `has_caio_enterprise`, `evidence` (press, both, or blank), notes, `read_note` |
| `appointed.csv` | 24 | The enterprise-wide roster, with the sentence quoted from the page that was fetched |
| `combined.csv` | 30 | Enterprise-wide titles that fuse AI with a second remit, same columns |
| `former.csv` | 12 | Companies that had the role and no longer do, with what the source says |
| `repeaters.csv` | 9 | Pages that repeat the 43% / 94 / 40% figures, with page type, what each attributes them to, and whether it links rework.com |
| `figures_in_circulation.csv` | 10 | Every Chief AI Officer adoption figure found, with the population it measures |
| `changes_2026-09-16.csv` | 39 | Every classification, source and date change made in the 16 September 2026 rebuild |
| `changes_2026-09-17.csv` | 61 | Every change made in the 17 September 2026 reading pass, the list cross-checks and the second and third reviews, with the old value, the new value, the reason and, on 51 of the 61 rows, the URL |
| `METHOD.md` | | Queries, the reading pass, classification rules, fetching, limitations |

`query_pass` says which of the three query passes found the evidence: `caio` for
the two Chief AI Officer phrases, `combined` for the four fused-title phrases,
`brand` for the re-query under a company's common name, `linkedin` for the three
rows whose only claim is a profile title (Fastenal, Paccar, WESCO International),
and `reading` for the one row the reading pass found on a company page after no
query pass returned it.

`read_note` is one line per company recording what a person saw in that company's
cached results. All 500 rows carry one.

## Licence

CC BY 4.0. Attribution: ProfessionLens.