OpenAI to test ads in ChatGPT for US free, Go users

OpenAI plans to test ads in ChatGPT for select US users, including those on the free version and the new US$8-a-month “Go” plan.

The company aims to diversify revenue sources amid high operational costs, with no plans to show ads on sensitive topics or to users under 18.

OpenAI has stated that ads will not influence ChatGPT’s responses, nor will conversations be shared with advertisers.

The move marks a shift from the company’s previous stance, which was cautious about advertising.

The company’s CEO of applications, Fidji Simo, emphasized the importance of maintaining trust and objectivity in ChatGPT’s responses as it explores new revenue models.

🔗 Source: Bloomberg

🧠 Food for thought

Implications, context, and why it matters.

Pin down what data powers ChatGPT ads and how privacy is enforced

Clarify whether targeting uses chat content (parsing messages for contextual targeting).
Explain how it verifies age gating (age verification and filtering) for sensitive topics for users under 18.
Verify whether bottom-of-response ads stay separate from answers, since it says ads will not affect answers while it subsidizes free users and the $8-a-month Go tier.
Detail which privacy-preserving pipes it will use (clean rooms and cohort/context Application Programming Interfaces (APIs)) 1. Clean rooms are secure spaces for matching and measurement without sharing raw user data. Cohort or context APIs send grouped or contextual signals instead of personal identifiers.
Confirm whether measurement relies on clean rooms for frequency capping and attribution 2. Frequency capping limits how often a user sees an ad, and attribution ties outcomes to ad exposure.
Assess how OpenAI will measure its 800 million weekly users under IAB Tech Lab (the Interactive Advertising Bureau technical standards group) clean room standards 3.

Build the privacy stack for ads in AI chat interfaces

Advertising technology providers (demand-side platforms or ad networks) prioritize clean room integrations with cohort/context APIs to enable targeting and measurement in chat without exposing data 1.
Measurement startups offer clean room-based frequency capping and multi-party attribution (cross-partner crediting), plus reconciliation for conversational sessions 2.
Contextual engine builders (teams that classify content plus environment signals) deliver real-time signals for chat as a cookie-free option 4.
Identity-light analytics vendors (those with limited access to login-based identifiers) invest in probabilistic conversion modeling when deterministic IDs (stable, user-specific identifiers) are constrained 5.
Investors track budget shifts toward AI-driven search plus chat that fund data pipes and tooling 3.
Platforms with first-party data use clean rooms to collaborate while monetizing product queries in chat 6.
Enterprise software vendors (customer service or commerce platforms) package chatbot Q&A with clean-room measurement and privacy-safe audience creation to support sponsored product slots in chat 2.