Skip to main content

Honest advertising in the AI layer.

A position. Not a feature.

We built an ad network. We think ads are okay. We also think a lot of the ads that are going to be served inside AI outputs over the next five years are going to be bad — bad for users, bad for the apps that carry them, bad for the networks that serve them. This is what we believe about what advertising inside AI should mean, written down so we can be held to it and so other networks can choose whether to disagree in public.

Principle 1 — Ads inside AI must be labeled.

Every Surface Surfacedd serves is labeled as sponsored. The label is not subtle. It is not a footer. It is inside the Surface itself, where the user reads, sees, or hears the content. A developer cannot hide it by choosing a CSS class that renders it invisible; a paint of the Surface to the user always includes the disclosure.

The reason isn’t decorative. Users trust AI output as authoritative. They treat a chatbot’s answer as a synthesis of what’s true, not a sponsored arrangement. Undisclosed sponsorship in that context doesn’t just bend a commercial arrangement — it weaponizes a trust gradient the medium was built on. Disclosure is how we stay inside the bounds of what the user came to the product for. It is the cheapest integrity-preserving move available, and it has to be free of design negotiation.

Principle 2 — Ads must be placements, not edits.

Surfacedd never changes the organic answer an AI gives. If an AI recommends three products in its answer, Surfacedd does not pay to add a fourth to that list. If a developer wants to show a sponsor, it goes in its own Surface, separate from the answer. The Surface sits after or beside the organic response. It does not edit it.

The distinction between placement and manipulation is the difference between advertising and propaganda. Placements are transparent — the user sees a separate unit and can attribute the message to a sponsor. Manipulation is hidden — the user sees one answer and has no way to know whether it was paid for. A network that starts down the manipulation road can’t come back; the incentive gradient always runs in one direction. We are committing publicly to staying on the placement side.

Principle 3 — Developers can leave.

Any developer integrating Surfacedd can remove us in one line of code, at any time, with no penalty, no cooldown, and no contract term. No exclusivity. No minimum-term lockup. No recovery fees if the developer pulls the SDK the day after installing it.

Contrast this with networks that require a 6-month commitment to access premium inventory, or that charge an exit fee, or that retain partial revenue for 90 days after a developer terminates the integration. Those contract shapes look normal in the broader ad ecosystem because they reflect a history in which switching networks meant re-engineering page templates. In the AI app world, integrations are lightweight and switching should be, too. The best signal of an honest ad network is how easy it is to turn off.

Principle 4 — We will not invent categories users didn’t ask for.

There are dark patterns emerging in the adjacent AEO / GEO / AI-visibility markets: sponsored “recommendations” disguised as organic answers, paid brand placements inside “neutral” advice, AEO-washing where a brand pays to appear authoritative rather than to appear sponsored. These patterns will find buyers. Advertisers will ask for them, and developers would take the money.

Surfacedd commits not to build them. We do not sell access to users’ private conversations. We do not sell a guaranteed mention inside an AI’s organic answer. We will not launch a product that promises a brand appears in a chatbot’s citation list in exchange for payment. If you are looking for those services, they exist elsewhere, and that elsewhere is not us.

Principle 5 — If ads stop being the right model, we’ll say so.

Ads work when nothing better does. In 2026, for most AI apps below a certain scale, nothing better does. That’s what we’re building for. But if a better economic model for AI apps shows up — usage-based billing that clears at fair prices, revenue-sharing between model providers and the apps that host them, a governance arrangement that transfers payments directly from users to developers without a network in the middle — we will tell developers about it before our competitors do. We are in the ad business because it is the business that fits right now, not because it is the business that has to fit forever.

Who we’re signing this for: developers who want to make money without being ashamed of how, advertisers who want to reach AI users without being deceptive, and users who want to know when they’re being sold to. If any of the five principles above stops being true of Surfacedd, the network has lost the right to that name. Hold us to it.

— The Surfacedd team, 2026

Questions, disagreements, counter-examples: [email protected].