Self-attestation and disclosure

In the absence of cryptographic proof, the most widely deployed provenance mechanism is self-declaration: asking the submitter to state how their submission was prepared. This is cheap and universal but relies on honesty, and its effectiveness depends heavily on form design.

Academic publisher AI disclosure policies (2024–2026). Nearly every major academic publisher now requires authors to disclose AI tool usage, with no publisher allowing AI to be listed as an author. Science bans AI-generated text outright; Springer Nature / Nature prohibits AI authorship but allows undisclosed copy-editing; IEEE requires acknowledgment-section disclosure; Elsevier, Wiley, Taylor & Francis, SAGE permit AI use with detailed, section-tied disclosure. Version information is now standard: "ChatGPT (GPT-4o, OpenAI, accessed January 2026)" with task description. See Enago, AI Usage Cards, and Thesify.

Legal profession AI disclosure (post-Mata v Avianca). In June 2023, Judge P. Kevin Castel (SDNY) sanctioned attorneys who submitted a brief containing fabricated ChatGPT citations under Rule 11. The case catalysed a wave of judicial disclosure requirements; Judge Brantley Starr (N.D. Tex.) and the Fifth Circuit now require certification that filings were either not AI-drafted or human-verified. The problem did not abate: a research database tracking court decisions on AI-fabricated content had catalogued over 1,200 such cases worldwide by early 2026. See the AI Hallucination Cases Database (Damien Charlotin), the AI Court Disclosure Map 2026, Mata v Avianca, and Spellbook.

The Artificial Intelligence Disclosure (AID) Framework. An emerging structured framework featuring a self-assessment rubric, checkbox declarations with toggle descriptions, and version tracking (tool, version, access date, task). See the AI Declaration Statement template and Three Disclaimers for Safe Disclosure.

Conflict-of-interest declarations in government and parliamentary contexts. The UK Parliamentary Register of Members' Financial Interests requires registration within 28 days, with API access under the Open Parliament Licence. The UK's Procurement Pathway provides a standardized conflict-of-interest declaration form, Australia's Department of Finance maintains a clausebank template, and the UNODC documents international good practice. See UK Parliament, UK Procurement Pathway, Australian Department of Finance, and UNODC.

High transferability for the mechanism; uncertain effectiveness. Self-attestation is the lowest-barrier provenance signal: no special technology, no identity infrastructure, no changes to authoring tools. The academic and legal precedents show that structured disclosure forms specifying tool, version, task, and extent produce more useful information than binary checkboxes.

A government intake form could include a structured "preparation method" section: whether the submission was drafted personally, by someone else, or with AI assistance; if AI-assisted, which tools and for which parts; and whether it is made on behalf of an organization. The academic model of requiring this in a specific section translates to a structured field at submission time rather than text buried in the submission.

The critical limitation is that self-attestation without consequences is performative. The legal model works because false certification triggers professional sanctions; academic disclosure works because misconduct findings damage careers. Government consultations typically lack equivalent enforcement, so the declaration needs pairing with a way to bind it to a verified identity, or with meaningful consequences for a false declaration, before it carries weight.