Sponsored AI responses.

A sponsored AI response is a disclosed sponsor unit rendered alongside a chatbot’s organic answer. The user asks a question. The agent generates its reply. A labeled Surface appears next to that reply with one sponsor. Two separate outputs, one user moment. The sponsor pays for the Surface. The agent does not get paid to say anything specific in its own reply.

The term has started to mean several different things across the industry, so the definition matters. Some vendors use “sponsored response” to describe sponsor-influenced answers where the agent is trained or prompted to mention a brand inside its own output. That is a different product, with a different set of consumer protection issues, and Surfacedd does not sell it. What we sell is a separate unit in a separate visual container with its own label, running on its own auction.

The distinction is not a detail. It is the entire argument for why the format is sustainable. A user can tell the difference between the agent’s answer and a sponsor next to it because the Surface is structurally separated, every time, in every app, with a persistent label. The user knows where the agent ends and the sponsor begins. That knowledge is what makes the organic answer trustworthy and the sponsor Surface payable. Break the separation, and both sides of the economy collapse together. See the glossary definition for the formal term and honest AI advertising for the position behind it.

Disclosure runs at three layers. Each layer is independent of the others so that disclosure survives partial failure.

Layer one is the visual container. The Surface has its own border, its own background treatment, and its own spacing from the organic answer. The container does not share styling with agent output. A user scanning the screen sees two separate blocks before reading either.

Layer two is the sponsor label. The word appears above the headline, in the reading order, as text. The label persists if the user customizes the app theme, switches to high contrast mode, or renders the app through an assistive technology stack. A label that disappears when the user changes a setting is a label that does not survive production.

Layer three is the API payload. Every sponsored response Surface is served to the host app with an explicit sponsor flag in the API payload. Apps cannot accidentally render a sponsor Surface as organic content because the API tells the app what it is receiving. Apps that intentionally strip the flag are cut from the network the same week we detect it. The flag is auditable on the advertiser side and on the publisher side.

Together the three layers enforce a disclosure standard that does not depend on any one app’s design discipline. The disclosure is the product, not a policy layer on top of the product.

ChatGPT Ads is OpenAI’s first-party advertising product inside ChatGPT. It places sponsored content adjacent to answers ChatGPT generates. The format is similar in shape to sponsored AI responses on Surfacedd. The differences matter at campaign setup.

Surface scope. ChatGPT Ads runs inside ChatGPT only. Surfacedd runs across the third-party AI app network — chatbots, writing apps, research tools, customer support agents, category copilots. An advertiser who wants coverage of the whole AI layer runs both.

Audience overlap. ChatGPT has roughly 800 million weekly users. Third-party AI apps reach a partially overlapping but distinct audience that tends to spend time in specialized tools for specific workflows. The overlap is lower than most advertisers assume at the first look; the gap closes only if you run both networks.

Pricing mechanics. ChatGPT Ads uses a proprietary auction inside OpenAI’s own inventory. Surfacedd runs an open auction with published floors. Surfacedd reports Share of Placement across the network; ChatGPT Ads reports inside its own surface. Reconciliation between the two is manual today; we publish a methodology for combining the signals.

Targeting. Both systems target contextually. ChatGPT Ads uses its own signal set from the ChatGPT product. Surfacedd targets by context bucket, app inclusion list, and geography. Neither requires third-party cookies. Neither relies on persistent user identifiers.

Operational difference. ChatGPT Ads is one integration with one publisher. Surfacedd is one integration with a network. The operational load on your side is the same — one account, one creative asset set, one reporting dashboard per network — but the reach scope and the pricing mechanics differ. For the longer position on where Surfacedd fits alongside first-party AI ad products, see reach AI users.