Observatory · Q3 2026

The field, measured each quarter.

38%

of public web text sampled was likely machine-generated

Provenance index, up from 29%

1 in 4

deployed hiring models failed an intersectional audit

Fairness index, 212 audits

governments with a published AI governance charter

Civic index, up from 41

of frontier labs publish a model welfare policy

Welfare index, up from 3%

Methodology

All figures are derived from Orrery Institute primary research. Provenance index samples 10,000 documents per quarter from Common Crawl. Fairness index covers audits conducted or reviewed by Institute researchers. Civic index tracks published charters meeting minimum criteria. Welfare index covers labs with at least one frontier model above 70B parameters.