AI advertising disclosure: what ChatGPT's ad platform means for consumers and founders

An AI ad is not an ad you recognize as an ad. That's the disclosure problem in one sentence, and it's why ChatGPT's move into advertising this spring matters far beyond what brands are paying per click.

The Trust Transfer Problem is what happens when advertising expands into a new channel before audiences have learned to discount it. It happened with TV, with search, with influencer content, and it's happening again now. The difference is that AI responses feel like advice, not like ads. That's a meaningfully different kind of trust to trade on.

What the format actually looks like

ChatGPT's current ad format surfaces inline with responses. When you ask a question with commercial intent, a "Sponsored" label appears next to a result. The ad content is integrated into the answer, not displayed separately in a sidebar or a clearly demarcated zone.

The launch partners are Target, Adobe, Williams-Sonoma, Albertsons, Georgia-Pacific, DSW, and The Knot. The agency holding groups with early access are Dentsu, Omnicom, Publicis, and WPP. Pricing started at roughly $60 CPM, moved to $25 CPM in the second cohort, and has since shifted toward CPC models in the $3-5 range.

The disclosure gap

OpenAI's current approach labels ads as "Sponsored." That's the same word Google has used since 2013. The problem is that 49% of users don't notice it.

This isn't because users are inattentive. It's because the format doesn't pattern-match to what people recognize as an ad. A Google result has a visual structure that signals "this is a paid result." A ChatGPT response doesn't; it reads as a recommendation from a knowledgeable source.

The FTC has general disclosure requirements for sponsored content, but no AI-specific guidance. The ASA in the UK has started flagging AI-generated content that lacks adequate disclosure, but hasn't produced formal rules for AI search. Both regulators are watching; neither has acted.

What this means for founders building on AI channels

If you're building distribution on AI-native channels, the absence of clear ad formatting is temporarily in your favor and structurally risky.

High conversion rates in a low-scrutiny environment don't sustain. Audiences learn. Every new advertising channel follows the same arc: early adopters get outsized returns because trust is high and skepticism is low, then disclosure pressure forces format changes, then performance normalizes.

The influencer marketing channel ran this exact playbook from 2015 to 2021. Early practitioners got extraordinary returns. Then the FTC's 2019 endorsement guidance and the 2021 enforcement wave changed the disclosure standard. Now #ad is table stakes.

The founders who built durable influencer channels weren't the ones who maximized short-term reach before the crackdown; they were the ones who built genuine audience trust that survived the format change. The same logic applies to AI advertising.

The regulatory picture right now

The FTC's existing endorsement guides apply broadly to sponsored content, including AI-generated recommendations, but the agency hasn't issued AI-specific enforcement guidance. The ASA (UK) updated its influencer guidelines in 2023 to include AI-generated content, and has indicated it's monitoring AI search advertising, but again: no specific rules yet.

The gap creates short-term opportunity and long-term exposure. Brands using ChatGPT's ad platform today are operating in a regulatory window, not a regulatory clear zone. That window will close; the only question is the timeline.

The Trust Transfer Problem, defined

The Trust Transfer Problem is the mechanism by which advertising uses a channel's ambient trust before that trust has been discounted for advertising. A new medium is trusted because it's new; advertising exploits that trust; audiences learn to discount it; the medium matures.

The loop always closes. The question is whether you're building a brand that benefits from the gap or one that depends on it.

For founders: the AI advertising channel is not a sustainable moat. It's a window. What you build inside that window, in terms of genuine audience relationship, is what persists after it closes.