Watermarking (AI content)
Embedding a statistically-detectable signal into generated content so downstream verifiers can identify it as AI-produced. Different implementations for text (token-level bias) and images (imperceptible frequency-domain patterns). Increasingly part of provider defaults; complements C2PA content provenance.
Related terms
- C2PA
Coalition for Content Provenance and Authenticity — an open standard for cryptographically signing the origin and edit history of digital content (images, video, audio, PDF). Adopted by Adobe, Microsoft, OpenAI, Google, TruePic. Foundational for AI-generated content labelling under the EU AI Act.
- EU AI Act
European Union regulation (effective 2024, phased through 2027) that classifies AI systems into risk tiers — unacceptable (banned), high-risk (conformity assessment required), limited-risk (transparency obligations), and minimal — and imposes proportionate obligations. Applies extraterritorially to any provider placing AI on the EU market.
Related on this site
Where this fits
Watermarking (AI content) is part of the Evaluation & Governance vocabulary used in the Generative AI Maturity Framework. See the full glossary for the complete set of 149 defined terms, or take the free maturity assessment to see where your organisation stands.