AI Readiness benchmark for AI / Machine Learning

Across 477 audited domains we have classified as AI / Machine Learning, the median AI Readiness score is 49 out of 100. That is 10 points above the median for all 30,738 audited domains in the corpus (39).

How the scores are distributed

Half of the 477 audited AI / Machine Learning domains score between 42 and 56. The corpus-wide interquartile range, across all 30,738 audited domains, is 32 to 45.

Which factors this industry fails most often

The rubric scores 19 factors. A factor counts as failed here when a site scores below 50 out of 100 on it. The last column is how many audits each rate was calculated from. Media Accessibility and Question Coverage & FAQ Presence are not scored on every kind of page: a rankings table or a paginated index has no question to answer and no image to caption. Citation-Worthiness, E-E-A-T Signals and Question Coverage & FAQ Presence lean on English-language patterns, so on a page written in another language they are scored out of a smaller total. Those rows therefore leave out every audit of a page the factor was not scored on the same scale, and every audit that did not record enough about the page for us to tell. Every other row counts every audit in this industry that scored the factor.

Where AI / Machine Learning diverges from everyone else

Relative to the whole corpus, this industry is distinctively weak on Content Freshness (+23 points), Structured Data Quality (+3 points).

It is distinctively strong on Trust & Brand Signals (-44 points), Answer Box Potential (-43 points), Schema Markup Completeness (-43 points).

Content Freshness: fails on 72% of the 477 AI / Machine Learning audits it was measured on

Freshness is the only unanimous causal gatekeeper across engines in the 2026 evidence base, and AI-cited content skews measurably fresher at web scale. Google AI Overviews is the documented exception.

Structured Data Quality: fails on 99% of the 477 AI / Machine Learning audits it was measured on

Rescoped to extractable data structures in the HTML itself (tables and lists), which measurably improve machine extraction. This is distinct from schema.org markup, which shows no AI-citation lift in controlled tests.

Audited AI / Machine Learning sites with a public report card

Each of these sites has a public AI Readiness report card from a real audit. Scores are point-in-time; the card shows what was measured and when.

Browse every audited site in the public audit archive.

See the full ranked list of 50 AI / Machine Learning sites →

How to read this

• What the sample is. The 477 domains we have audited and classified as AI / Machine Learning. That is a sample of the industry, not a census: it is the set of sites our crawler has reached and scored, and it carries whatever bias that selection has.
• Rubric version 2026-08-13. Every figure here comes from audits scored on this single rubric generation. Scores from different generations measure different things, so they are never mixed. Compatible generations: 2026-08-13.
• Why there are no company names. These are aggregates. We do not publish which domains sit at the bottom of a distribution, and the queries behind this page cannot return a domain name.
• Publishing floor. We publish an industry only once it reaches 100 audited domains, and a factor's failure rate only once 30 audits in the industry count toward that factor. Below either floor the number would move with a handful of sites, so we withhold it.
• What the audit can and cannot see. This audit measures citation readiness: how easily AI engines can crawl, extract, and quote these pages. It cannot see the off-site brand footprint (mentions, reviews, third-party coverage) that gates whether AI engines retrieve a brand at all; Kuroma's visibility scans and off-site signals track that side. Read the full methodology.

Other industry benchmarks

All industry benchmarks