AI
Public AI interactions
Chat, intake, service, recommendation, or other visitor-facing AI-assisted tools that are observable on the site.
Public-facing AI transparency review
Can your website show where its AI came from?
Run a free SiteTrust AI Transparency Scan. See the public AI interactions, disclosures, trackers, and supported proof-of-origin signals a visitor can observe on your website today.
No installationNo site accessPublic observationDated results
The question is public
Visitors can already see AI on your website. The question is whether they can understand what it is, how it is disclosed, and what proof of origin is available.
Not a policy review
Not what the site says. What the public page does.
An AI chat assistant on a contact page. An intake bot in a form. AI-generated media in a campaign. A support tool that answers questions. A disclosure written at launch, never updated when the site changed.
None of this requires access to a company's systems to see the first question: what does a visitor encounter on the live public website right now?
SiteTrust does not accuse a company of breaking the law. It produces a public measurement, on a date, with a stated method and scope.
?
Disclosure signals
Whether the public page explains that a visitor is interacting with AI or a system that uses AI.
◈
Proof-of-origin signals
Supported checks for available Content Credentials, C2PA provenance metadata, and other detectable public signals.
◎
Tracking context
Relevant public tracker and consent behavior, including what loads and what a visitor can observe around the interaction.
The AI Transparency Scan
Five public signals that affect website trust.
The scan is designed to show observable technical and disclosure evidence. It does not infer private model use, legal status, or authorship from a missing signal.
AI interaction inventory
Where visitors encounter observable AI-assisted chat, intake, support, or recommendation experiences.
AI disclosure presence
Whether a public page provides a clear explanation that a visitor is interacting with AI or an AI-supported system.
Content Credentials
Whether supported public assets carry available C2PA/Content Credentials provenance metadata.
Watermark and provenance signals
Whether supported assets show detectable public provenance or watermark-related signals within the scan's documented methods.
Consent and tracking context
Whether relevant public tracking behavior around the website experience is observable before or after a visitor makes a privacy choice.
A record, not a guess
Each result is tied to a date, stated public-facing scope, and a path to the full SiteTrust evidence report.
Scope matters
Be precise about what a scan can and cannot say.
What SiteTrust can report
- Publicly observable AI interactions and disclosure signals.
- Available provenance metadata on supported public assets.
- Detectable signals within the stated watermark/provenance method.
- Public tracking and consent context within the scan scope.
- What SiteTrust observed, when it observed it, and how the result is categorized.
What SiteTrust will not claim
- That every AI-generated asset can be detected.
- That missing metadata proves content is human-created.
- That a disclosure or scan result alone establishes legal compliance.
- That SiteTrust has inspected private models, prompts, training data, or internal systems without access.
- That a provenance signal proves content is true, safe, or reliable.
The SiteTrust path
Scan. Understand. Fix. Prove. Keep watching.
1
Scan a public domain
Start with a public website. No installation, account access, or code change is required for the initial observation.
2
Receive a dated summary
See high-level findings and observable signal categories from the scan.
3
Review the evidence
Verify and Audit provide evidence, context, a fix path, and the applicable SiteTrust standard.
4
Publish proof
Qualifying sites receive a badge and a public SiteTrust registry record that visitors can inspect.
5
Review on schedule
Trust is maintained through recurring live scans and reviews as websites, vendors, and public AI experiences change.
The badge is the front door
The registry record is the proof.
A SiteTrust badge should open a record anyone can check. It should show what was reviewed, the date of review, the current status, the stated scope, and when the next scheduled review is due.
For AI transparency, the record can distinguish public disclosure, visible AI interactions, and available proof-of-origin signals without suggesting that SiteTrust has made a universal claim about every asset or internal system.
View the public registryCurrent review status
AI transparency review
Public AI interaction disclosurePresent
Supported Content CredentialsObserved on selected assets
Provenance-signal scopePublic assets only
Last public scanAugust 2026
Next scheduled reviewNovember 2026
This example shows the type of scoped, dated evidence a registry record can provide.
Built for reputation-dependent websites
When your website is part of your credibility, public evidence matters.
Professional services
Law, accounting, insurance, wealth, consulting, recruiting, healthcare-adjacent, and other firms whose website is part of a prospect's due diligence.
AI-enabled companies
Businesses with public chat, support assistants, AI-led intake, AI content, or automated decision-support experiences.
Agencies and advisers
Teams that build, manage, inherit, or advise on many client sites, and need a repeatable, evidence-based service line.
Questions the page should answer
Clear scope builds trust.
Use answers like these in the final production page, scan results, sales deck, and registry methodology.
Does the scan need access to our website or systems?
No for the initial public scan. It observes publicly accessible website behavior in a manner similar to what a visitor's browser can encounter. It does not install code, access private systems, or change the site.
Can SiteTrust tell whether every piece of content was made by AI?
No. The scan reports supported, publicly observable signals and available metadata within its stated methods. A missing watermark or Content Credential does not prove that content was created by a person.
Does a positive result mean our website is legally compliant?
No. The result is technical and public-facing evidence within the scan scope. Legal compliance depends on the facts, applicable jurisdictions, disclosures, implementation, data practices, contracts, and other considerations.
What happens after a scan identifies a finding?
The free scan provides a summary. SiteTrust Verify and Audit can provide the evidence report, remediation path, ongoing scans, verification, and eligibility for the badge and public registry record, subject to the applicable program requirements.
Why should an agency use this?
Agencies can verify their own domain, see the same report their clients will see, and use SiteTrust as a repeatable website-trust service. The SiteTrust partner program can support referrals and implementation work under the finalized partner terms.
Start with what is public
See what your visitors can see about AI on your website.
Run a free, dated public scan. Then decide whether you want the evidence, remediation path, recurring review, badge, and registry record behind it.
Run your free AI scan