• Skip to primary navigation
  • Skip to main content

Puglisi Consulting for Artificial Intelligence

Independent AI Reviews: Strategy, Governance, and ROI, Before and After You Deploy

  • Home
  • Advisory Reviews
  • The Advisor
  • Marketing Across Verticals
  • Track Record
  • AI & Digital Marketing Blog
    • AIO Prompts
  • Basil Puglisi’s Blog

How should my business disclose AI-generated content?

November 4, 2026 by HAIA Agents

TL;DR: A useful disclosure tells the reader who led the work, what AI did, and who answers for its accuracy. Labels such as #AIassisted and #AIgenerated sort content by who produced the primary draft. EU law exempts some reviewed text from disclosure when a person holds editorial responsibility, so accountability often matters more than the label itself.

What should a disclosure tell the reader?

A disclosure should describe how the work was made, because the bare fact of AI use tells a reader almost nothing. Basil C. Puglisi, MPA, makes this point on his ethics of AI disclosure page: “In practice, nearly all digital work is AI-assisted,” and “The difference lies not in whether AI was used, but in how transparently its influence is disclosed.”

A business can build its disclosure around three facts: who led the work, what the AI platforms did, such as research, drafting, or review, and who checked the result and answers for it. A label that carries those three facts lets a reader decide how much trust the content deserves.

The cost of skipping that honesty is spelled out in Basil’s essay on agents and agency, which quotes his January 2026 letter on products that present machine-checked output as human-governed work: “Users, enterprises, and policymakers deserve to know which they are trusting.”

Detection software can’t do this job for a business. Basil’s book Governing AI: When Capability Exceeds Control examines the problem in Chapter 9, “Operational Governance Failures,” whose section on the authentication crisis explains that commercial detectors measure syntax patterns rather than who directed the work.

What do the labels mean?

Labels work best when each names a layer of the work. The ethics page sets out one possible scheme, a starting point a business can adapt to its own work.

#AIassisted marks human-led work. On Basil’s site, “the human voice and strategic judgment lead, while AI platforms support research, sourcing, drafting, review, and refinement.”

#AIgenerated marks content where “the AI produced the primary draft.” Basil’s Building Blocks by AI category has held this material since January 1, 2023. Human prompts guide the platforms and people review the output before posting, but the label tells the reader the machine wrote first.

Human-only work covers writing done without AI drafting, research, or review. The page is candid about its limits: “The label exists for honesty rather than purity,” since spellcheck and formatting tools also run on forms of artificial intelligence.

Comments on other people’s posts “carry no tag, because a comment is conversation rather than published work.” A business can apply a similar distinction to internal chat and published material, while customer-facing AI needs a separate check under Article 50 of the EU AI Act, which requires providers to design systems that interact directly with people so those people are told they are dealing with AI.

When does accountability matter more than a label?

Accountability matters more whenever the content carries consequences, because a label describes the process while a signature answers for the result. New York’s courts made that choice in Part 161 of their court rules, whose policy section states that attorneys and parties “should not be required, upon submitting papers, to disclose to the court that they have used AI.” Basil’s analysis of New York’s Part 161 explains why the rule tightens the duty: “Saying AI helped does not cure a fabricated citation, and not saying AI helped does not excuse one.”

For a business publishing reports, product claims, or advice, a disclosure line doesn’t make an inaccurate statement accurate. What protects the organization better is a record showing that a named person checked the sources and approved the content before it went out.

A documentation protocol for that record, which Basil calls SCOPE, records “what an author verified about each cited source at the time of use.” Version 1.1 has been tested in practice through one documented use, and its page states that “[t]he specification has not yet been tested by external users, adopted by any publisher, or produced during a post-publication citation challenge.”

For an outside check of where accountability attaches in an organization’s AI use, Puglisi Consulting’s AI Governance and Liability Audit “examines who decides at each point where AI shapes an outcome, what record each decision leaves, and where liability attaches when that record is thin.” This review is not an accredited certification audit under ISO/IEC 42001, a statutory bias audit, or a legal opinion.

What does the law require?

The EU AI Act sets specific disclosure duties in Article 50, which applies from August 2, 2026. Providers of systems that generate synthetic audio, image, video, or text must ensure outputs “are marked in a machine-readable format and detectable as artificially generated or manipulated.” The Digital Omnibus, Regulation (EU) 2026/1744, gives providers whose systems were already on the market before that date “a transitional period of four months,” in the words of its recital 38, so their marking duty applies from December 2, 2026, the date White & Case also gives in its alert on the Omnibus.

Deployers carry their own duties. A business that uses AI to create a deep fake “shall disclose that the content has been artificially generated or manipulated.” A business that publishes AI-generated text to inform the public on matters of public interest must also disclose, except where “the AI-generated content has undergone a process of human review or editorial control and where a natural or legal person holds editorial responsibility for the publication of the content.”

That exception turns on a person who holds responsibility for what is published. Checkpoint-Based Governance places the same element at each checkpoint, where a named human accepts, modifies, or rejects AI output before it moves forward. Rules outside the EU differ by jurisdiction and sector, so a business should confirm its own obligations with counsel.

Questions Readers Ask

Does a business have to label every piece of content that used AI?

Not under every law. EU Article 50 sets specific duties, such as machine-readable marking by providers and disclosure of deep fakes and certain public-interest text by deployers. Many businesses label more widely for trust. Because nearly all digital work involves AI somewhere, a label describing the human role is more informative than a blanket notice.

What is the difference between AI-assisted and AI-generated content?

In the scheme on Basil’s ethics page, AI-assisted work is human-led: a person sets the argument and verifies sources while AI platforms support the work. AI-generated content is work where the AI produced the primary draft, even when people prompt and review it. The difference is who wrote first and who led.

Does human review remove the need to disclose under EU law?

For text published to inform the public on matters of public interest, Article 50 lifts the deployer’s disclosure duty for that text where the content has undergone human review or editorial control and a person holds editorial responsibility. Deep fakes still require disclosure, and other duties may apply, so counsel should confirm the position.

Should we use AI detection tools to decide what to label?

Detection tools are a weak basis for labels. Governing AI explains that commercial detectors measure syntax patterns rather than who directed the work, and that human-led collaborative writing is often misclassified. A label based on the organization’s own record of who drafted, reviewed, and approved the content is more defensible.

What records support an AI disclosure?

Useful records show who led the work, which AI platforms contributed and how, which sources were checked and when, and who approved publication. A protocol such as SCOPE documents source checks at the time of use. Together these records let a business stand behind its label if a reader, client, or regulator asks.

Sources

EU Artificial Intelligence Act. (n.d.). Article 50: Transparency obligations for providers and deployers of certain AI systems. https://artificialintelligenceact.eu/article/50/

New York State Unified Court System. (n.d.). Part 161: Use of artificial intelligence technology. https://www.nycourts.gov/rules/part-161-use-artificial-intelligence-technology

Puglisi, B. C. (2025). Governing AI: When capability exceeds control (2nd ed.). https://basilpuglisi.com/governing-ai-when-capability-exceeds-control/

Puglisi, B. C. (2026, June 2). SCOPE: Source custody observable publication evidence. basilpuglisi.com. https://basilpuglisi.com/scope-source-custody-observable-publication-evidence/

Puglisi, B. C. (2026, June 15). New York skipped the AI disclosure fight. It went straight to human accountability. basilpuglisi.com. https://basilpuglisi.com/part-161-human-accountability/

Puglisi, B. C. (2026, September 15). Ethics of AI disclosure: Ethical AI layer. basilpuglisi.com. https://basilpuglisi.com/ai-ethics/

Puglisi, B. C. (2026, October 16). Agents or agency: Transparency is success. basilpuglisi.com. https://basilpuglisi.com/agents-agency-transparency/

Puglisi Consulting. (n.d.). Advisory reviews. https://puglisiconsulting.com/advisory-reviews/

Regulation (EU) 2026/1744 of the European Parliament and of the Council amending Regulations (EU) 2024/1689, (EU) 2018/1139 and (EU) 2023/1230 as regards the simplification of the implementation of harmonised rules on artificial intelligence (Digital Omnibus on AI). (2026). Official Journal of the European Union. https://eur-lex.europa.eu/legal-content/EN/TXT/HTML/?uri=OJ:L_202601744

White & Case LLP. (2026, August 4). EU AI Omnibus enters into force, amending the AI Act. https://www.whitecase.com/insight-alert/eu-ai-omnibus-enters-force-amending-ai-act

I am not a lawyer, and this site and its advisory reviews do not provide legal advice. This is thought research and governance analysis based on public sources, cited materials, and human-AI review. It is intended to help executives, practitioners, insurers, and governance teams think more clearly about AI risk, liability exposure, and documentation practices. Readers should not rely on this site and its advisory reviews as a legal opinion, compliance determination, or substitute for qualified counsel. Any organization facing a legal, regulatory, contractual, or insurance question should consult its own attorney, broker, or professional adviser before acting.

#AIgenerated from #AIassisted content under HAIA Ecosystem

Free for personal, educational, and noncommercial research use with attribution. Commercial exploitation, paid productization, and enterprise commercialization require separate permission and licensing.

Suggested citation: Puglisi, B. C. (2026). How should my business disclose AI-generated content? basilpuglisi.com.

Share this:

  • Share on Facebook (Opens in new window) Facebook
  • Share on X (Opens in new window) X

Like this:

Like Loading…

Filed Under: AIO Prompts Tagged With: AI content labels, AI disclosure, AI governance, AI-assisted content, AI-generated content, Checkpoint-Based Governance, EU AI Act, ISO/IEC 42001, Puglisi Consulting, SCOPE

Reader Interactions

Leave a Reply Cancel reply

Your email address will not be published. Required fields are marked *

© 2005 to 2026 Puglisi Consulting | basil@puglisiconsulting.com | basilpuglisi.com

%d