Skip to main content
v2026.11,986 entries · CC-BY 4.0
CASRAICASRAI Working GroupsHelp write the standards you have to implement24 working groups propose, review and ratify the terms CASRAI publishes, one per dictionary domain. Community and observer seats start at about two hours a month; chairs and co-chairs give 6–8 hours a month. Applications are reviewed.Apply to join a group CASRAI · Volunteer seats
Dictionary termTrack AProposedv2026.1

AI Transparency Marker

An AI transparency marker is a visible label, watermark, or embedded machine-readable metadata tag attached to a piece of content — text, image, audio, or video — to disclose that it was generated or substantially modified by an artificial intelligence system.

The concept spans two related but distinct mechanisms: human-readable disclosure (a statement in a manuscript, caption, or figure legend that AI assistance was used) and technical marking embedded directly in the file itself (for example, a C2PA Content Credential or an invisible watermark) that software can detect even if the disclosure text is stripped or ignored.

ByCASRAI Editorial Board
· Last updated 5 Sept 2026

Working group for this domain: Generative AI use and disclosure working group

Share this

Ask CASRAI · full access

Ask about AI Transparency Marker

Full access: large monthly usage, up to 10 source passages behind each answer, and MCP and API access for Claude Code, Codex and other CLIs. Cancel any time.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Editorial commentary

An AI transparency marker is a visible label, watermark, or embedded machine-readable metadata tag attached to a piece of content to disclose that it was generated or substantially modified by an AI system. It sits alongside, but is distinct from, a written disclosure statement — a marker is the technical mechanism that lets software, not just a human reader, detect that content is AI-generated.

Two regulatory pushes behind the concept

The practical importance of AI transparency markers in research and publishing workflows is driven by two separate but related regulatory developments:

  • The EU AI Act’s Article 50(2) requires providers of AI systems that generate synthetic audio, image, video, or text content to mark the output in a machine-readable format that is detectable as artificially generated or manipulated. Under the 2026 Digital Omnibus, this obligation’s deadline moved from 2 August 2026 to 2 December 2026.
  • US state law, most notably California’s AI Transparency Act (SB 942, as amended by AB 853), requires large generative-AI providers to offer both a “manifest” (visible) and “latent” (embedded, machine-readable) disclosure option for AI-generated image, video, and audio content.

How it’s implemented technically

The leading technical mechanism for embedding a machine-readable marker is C2PA content provenance (Content Credentials) — a cryptographically signed metadata record chained to the file. Simpler invisible-watermarking approaches also exist, embedding a detectable signal directly in generated pixels or audio without a full provenance chain.

Why this matters for research administration

Journals and funders increasingly need to distinguish narrative AI-use disclosure (an author’s statement in the methods section) from technical, machine-checkable evidence that content was AI-generated. A submission system that can programmatically check an embedded marker doesn’t have to rely solely on an author’s self-report — the two mechanisms are complementary, and policies are increasingly asking for both.

Related terms

Examples

Worked examples

  • Is an instance

    A generative-AI image or figure-creation tool embeds a machine-readable marker in every output file, so a journal's production system or a reviewer's tooling can flag AI-generated figures automatically, independent of whether the author separately wrote a disclosure sentence.

  • Is an instance

    A funder-mandated data-visualization pipeline attaches a signed provenance record to each AI-assisted chart it exports, giving downstream reviewers an inspectable, tamper-evident trail of what the AI touched.

Counter-examples

Looks similar, but isn't

  • Not an instance

    A plain-text disclosure sentence alone ("Portions of this manuscript were drafted with the assistance of [tool]") is a disclosure statement, not an AI transparency marker in the narrower technical sense used here, unless it is paired with an embedded, machine-detectable tag. See Watermarking (AI output) and AI provenance for the narrative-disclosure and provenance-tracking sides of this practice respectively.

Machine-readable encodings

Use in your systems

JATS XML <kwd> element
xml
<kwd vocab="CASRAI Dictionary"
     vocab-identifier="https://casrai.org/dictionary/"
     vocab-term="AI Transparency Marker"
     vocab-term-identifier="https://casrai.org/dictionary/term/ai-transparency-marker">AI Transparency Marker</kwd>
Schema.org DefinedTerm (JSON-LD)
Schema.org DefinedTerm (JSON-LD)
json
{
  "@context": "https://schema.org",
  "@type": "DefinedTerm",
  "@id": "https://casrai.org/dictionary/term/ai-transparency-marker",
  "name": "AI Transparency Marker",
  "identifier": "https://casrai.org/dictionary/term/ai-transparency-marker",
  "description": "An AI transparency marker is a visible label, watermark, or embedded machine-readable metadata tag attached to a piece of content — text, image, audio, or video — to disclose that it was generated or substantially modified by an artificial intelligence system. The concept spans two related but distinct mechanisms: human-readable disclosure (a statement in a manuscript, caption, or figure legend that AI assistance was used) and technical marking embedded directly in the file itself (for example, a C2PA Content Credential or an invisible watermark) that software can detect even if the disclosure text is stripped or ignored.",
  "inDefinedTermSet": "https://casrai.org/dictionary/domain/genai-disclosure#set",
  "url": "https://casrai.org/dictionary/term/ai-transparency-marker",
  "license": "https://creativecommons.org/licenses/by/4.0/",
  "publisher": {
    "@id": "https://casrai.org/#organization"
  },
  "author": {
    "@id": "https://casrai.org/#editorial-team"
  },
  "datePublished": "2026-08-22T09:45:33Z",
  "dateModified": "2026-09-05T17:43:38Z",
  "inLanguage": "en",
  "isAccessibleForFree": true
}

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →