← All pages

AI Disclosure

Definition: A clause requiring that AI-generated or AI-assisted content or interactions be disclosed to users as such. Unlike most clauses in this wiki, it’s user-protective rather than a rights grab — tracked here because it recurs in comparable form across GenAI providers’ conduct policies. Risk level: Low — this clause protects users rather than harming them. It’s tracked for cross-provider comparison of transparency obligations; the real risk signal would be its absence, not its presence.

Note on provenance: This concept page covers the sources/GenGA/ (GenGA/Open Terms Archive) dataset, which has no pre-tagged JSONL — all tagging here is LLM-assigned by direct reading. See each linked platform page’s methodology note.

Platforms Using This Clause

Common Wording

Indicate that the content is AI-generated in a way no user could reasonably miss or misunderstand. — ChatGPT, Acceptable Use Policy, 2025-11-11

All consumer-facing chatbots, including any external-facing or interactive AI agent, must disclose to users that they are interacting with AI rather than a human. This disclosure must be provided at a minimum at the beginning of each chat session. — Claude.ai, Acceptable Use Policy, 2025-11-11

You shall clearly disclose to your end users that the Output content is generated by AI, and may contain errors or omissions and are for reference only. — DeepSeek, Developer Terms, 2025-11-11

Clearly indicate that the output content is generated by artificial intelligence, to alert the public to the synthetic nature of the content. — DeepSeek, Terms of Service, 2025-11-11

Create content that you subsequently claim was created solely by humans, or otherwise misrepresent the provenance of information or content; or provide chatbot services without disclosing to end users that they are interacting with AI (unless it is obvious from the context). — Perplexity, Acceptable Use Policy, 2025-11-11

You may not (i) publish any Output generated by the Services without clearly citing the Services, or (ii) misrepresent the source of any Output or the fact that it was generated by artificial intelligence. — Perplexity, Terms of Service, 2025-11-11

Respect guardrails and don’t mislead… Don’t mislead people as to the nature and source of Outputs, including images. You should be transparent and disclose your use of AI assistance and potential limitations, as applicable. — xAI, Acceptable Use Policy, 2025-11-11

Not complying with laws or regulations, including by: …Misleading others or not being transparent regarding your use of AI. — xAI, Terms of Service, 2025-11-11

Notes & Trends

Anthropic’s rule is the most mechanically specific: a session-start disclosure for “all consumer-facing chatbots,” plus a separate High-Risk Use Case disclosure rule. OpenAI’s and DeepSeek’s versions apply only to publishing/disseminating AI output, not to live chatbot interactions generally. DeepSeek is the only provider that ties the disclosure duty to a content-accuracy caveat in the same sentence (“may contain errors or omissions and are for reference only”) — making it read more like a liability disclaimer than a pure transparency rule. Microsoft Copilot’s version is the only one framed around technical watermarking (“content credentials”) rather than plain-text disclosure. Perplexity is the only provider to pair the disclosure duty with a separate, mandatory attribution requirement — it must “clearly cite the Services” when publishing Output, a dual transparency mechanism the others don’t have. xAI is internally inconsistent: its Acceptable Use Policy treats disclosure as a soft recommendation (“you should be transparent”), while its Terms of Service lists “not being transparent regarding your use of AI” as a flatly prohibited act — two different strength levels for the same obligation within one company’s documents. None of these providers cite the EU AI Act’s AI-content-labeling provisions as the source of this obligation; each frames it as its own policy choice, not a stated regulatory requirement.