Answer capsule
AI can generate editorial-looking advertising at high speed. FTC staff guidance keeps the CMO test grounded in whether people recognize the commercial nature before and after they interact.
What the source establishes
- The FTC guide supplements the Commission's Enforcement Policy Statement on Deceptively Formatted Advertisements with staff guidance for digital media.
- The guide says an advertisement can be deceptive when its format materially misleads consumers about its commercial nature, even if the underlying product claims are truthful.
- FTC staff directs advertisers to evaluate the ad as a whole, including appearance, similarity to surrounding non-advertising content, and the context in which people encounter it.
- The guide says it cannot cover every issue, provides general guidance rather than a safe harbor, and emphasizes that necessary disclosures must be clear, prominent, understandable, and preserved when content is republished.
Review the generated asset as an experience
A disclosure token stored in a campaign system does not establish the audience's net impression. Review the headline, visual style, synthetic presenter, publisher context, recommendation widget, thumbnail, audio, landing page, and path into the content together. Record what a reasonable member of the intended audience is likely to understand before clicking or listening. AI-assisted production increases the number of variants; it does not convert format review into a metadata check.
The accountable team should translate this point into a named workflow, affected population, source data, human owner, approval right, exception path, retained evidence, and review date. That translation is what separates an interesting AI development from a decision that can be governed and evaluated.
Preserve commercial identity through the supply chain
Native assets can move from a publisher page into search results, social posts, email, affiliate placements, content widgets, and automated localization. Attach the advertiser, commercial relationship, approved claim set, required disclosure, language, placement rules, and withdrawal owner to every distributable version. Test whether export, cropping, summarization, translation, or partner republication removes or weakens the disclosure. The source treats republication as part of the advertiser's responsibility, not as an excuse for lost context.
The accountable team should translate this point into a named workflow, affected population, source data, human owner, approval right, exception path, retained evidence, and review date. That translation is what separates an interesting AI development from a decision that can be governed and evaluated.
Measure comprehension, not label presence
The guide says effective disclosures must be noticed, processed, and understood. Test representative devices, placements, languages, visual abilities, audio settings, and entry paths. Measure whether people identify the content as advertising before engaging, not merely whether the word sponsored exists somewhere in the DOM or media file. Keep the test population, prompt, asset version, placement, result, and remediation. A platform's disclosure feature is documented capability; only the rendered audience experience can support the campaign decision.
The accountable team should translate this point into a named workflow, affected population, source data, human owner, approval right, exception path, retained evidence, and review date. That translation is what separates an interesting AI development from a decision that can be governed and evaluated.
Keep guidance and legal approval separate
The FTC page is longstanding staff guidance and expressly provides no safe harbor. It does not decide every AI-generated persona, publisher relationship, affiliate flow, product claim, or jurisdiction. Use it to make format, identity, and disclosure evidence visible, then route the actual campaign through qualified review. Preserve which conclusion came from the FTC source, which came from the brand's policy, which was observed in testing, and which remains unresolved.
The accountable team should translate this point into a named workflow, affected population, source data, human owner, approval right, exception path, retained evidence, and review date. That translation is what separates an interesting AI development from a decision that can be governed and evaluated.
Decision test
Ask whether the source changes the decision itself, the evidence required, the implementation sequence, or only the language used to describe an existing capability. Record which claims are directly supported, which are provider statements, which require an independent test, and which remain unknown. A source-linked review should make uncertainty easier to see, not bury it inside a blended score.
Questions to take into review
- What training, input, and output rights apply?
- Which review gates cover claims and brand expression?
- Which repository owns approved content?
- How are market and channel variations controlled?
- Can the asset's origin and edits be reconstructed?
- Which disclosures apply by market and context?
The publication supports research and executive decision preparation. It does not provide legal, financial, accounting, employment, clinical, cybersecurity, investment, procurement, or implementation advice.