Who Should Sign an AI Disclosure

An AI disclosure should usually be signed or approved by the person who can responsibly stand behind the final work. The goal is not to choose the person who touched the most software. The goal is to identify the person who reviewed the finished result and is authorized to approve it.

This page explains how organizations should think about that decision.

The signer should represent responsibility

In most cases, the signer should be the human decision-maker who reviewed the final work and has authority to approve its release. That may be the author, editor, project lead, manager, researcher, or another responsible reviewer depending on the context.

The signer is not necessarily the person who operated the AI tool. The more important question is who approved the final result.

Different organizations may choose different approvers

In a small organization, the signer may be the person who created and finalized the work. In a larger organization, the signer may be an editor, principal investigator, product owner, or department lead with authority to approve the deliverable.

The right answer depends on governance and the nature of the work, but the principle is the same: the signer should be able to defend the disclosure and the final output.

What the signer should know

The signer should understand what role AI played, what human review occurred, and what the disclosure is asserting. A signature should not be a ceremonial step added after the fact without understanding the underlying process.

If the signer cannot explain the disclosure, the process probably needs more review before publication.

Why this matters

Questions about AI use usually become questions about accountability. People want to know not only whether AI was used, but also who is taking responsibility for the final work. The signer answers that question.

That is why a disclosure should be tied to a real person with real approval authority.

How this relates to Provenance Disclosure

Provenance Disclosure is built around that idea of accountability. It helps organizations create a structured record that describes the role of automation while making clear who reviewed and approved the final result.

For related guidance, see AI Disclosure Statement and the documentation on signatories.

Create a Provenance Disclosure

If you are ready to document who reviewed and approved the final work, generate a Provenance Disclosure.