Search advertising largely monetized the query.

Search advertising largely monetized the query.

Conversational AI can monetize the formation of the decision.

That is a much more powerful position.

On 31 August 2026, OpenAI reported that ChatGPT Ads reached $1 billion in annualized revenue run rate in less than 200 days. On the same day, the European Commission designated ChatGPT a Very Large Online Search Engine under the Digital Services Act after it reported at least 45 million average monthly users in the EU.

OpenAI says ads are clearly labeled, separate from answers, and do not influence them. Advertisers do not receive private conversations. Those distinctions matter and should not be replaced by easy accusations.

OpenAI also says its ads system uses the context of the current conversation and, depending on country and settings, may use context from the user's broader ChatGPT experience.

No villain is required for the governance problem to exist.

The same interface can occupy at least two roles:

• help the user make a decision;
• select a commercial placement relevant to that decision.

Traditional search advertising was built largely around the query.

A conversational system may encounter the goal, budget, constraints, uncertainty, rejected alternatives, and the exact moment in which intention becomes a choice.

The valuable object is no longer only attention.

It is decision context.

On 2 August, I argued that the decisive AI question would increasingly become:

"What social role is it occupying in my life?"

Now that question is becoming operational.

A "Sponsored" label identifies an advertisement. It does not, by itself, prove the separation of roles behind the interface.

A mature evidence chain should eventually answer:

Which role was active?
Whose interest did it serve?
Which memory and inferences could it use?
Which policy governed it?
What information could leave the boundary?
Who was responsible, and how could the decision be challenged?

I call this role provenance.

We already ask which model ran and which data it accessed. Soon we will also need to know which social role acted, for which principal, and under what authority.

AI transparency begins with disclosure.

AI governance begins when the boundary between roles becomes inspectable.

The same voice may serve different interests.

Those interests must never be allowed to become invisible.

Source note:

OpenAI, "A milestone in expanding access to AI", 31 August 2026.

European Commission, "Commission designates ChatGPT, Reddit, Roblox under the Digital Services Act", 31 August 2026.

Earlier framework:

Ivan Kotov, "Separation and Composition of AI Social Roles, and the Custody of Memory and Interpretations", DOI: 10.5281/zenodo.21751985.

Claim ceiling:

These events make the role-separation problem operational. They do not prove that advertising influences ChatGPT's answers, and they do not validate any specific architecture by themselves.