General · Updated Jul 24, 2026
This week we pointed our own audit engine at atlascomplianceai.com — the same public-facing audit we run for clients. We publish AI-use disclosures in our footer and legal pages, we label AI-generated video on every platform, and we assumed we were covered. The audit disagreed.
The finding: our human-readable disclosure was in place, but the site carried no machine-readable AI-disclosure layer — no structured-data pointer to our disclosure page, no site-level disclosure metadata. A person reading the site would know exactly how we use AI. A crawler, a platform policy check, or another audit tool would find nothing. We scored it a medium-severity gap on our own rubric.
The fix shipped the same day: our Organization structured data now carries a publishingPrinciples pointer to our disclosure page, and every page serves a site-level ai-disclosure meta tag. Total change: six lines.
Why publish this? Because it is the honest version of the claim every compliance vendor makes. Audits work when they are allowed to find things — including on the auditor's own site. If your disclosure story lives only where humans can read it, the machine-readable half is probably missing. That is exactly the kind of gap we look for.
Business impact
If a compliance firm can ship a disclosure gap on its own site, so can you. Machine-readable disclosure layers are easy to miss because humans never see them.
Recommended action
Ask whether your AI-use disclosure exists in a form machines can read — metadata and structured data — not just a sentence in your footer.
Source: Atlas² AI Compliance