OpenAI Begins Offering Advertisers the Chance to Say Where They Would Rather Not Be

For most of the history of advertising, the question was where to place an ad. Now, increasingly, the question is where not to.

OpenAI, the San Francisco company whose ChatGPT service has become one of the more talked-about advertising venues of the moment, is testing what it calls negative targeting guidance options with a small group of advertisers, a spokeswoman, Taya Christianson, confirmed. The feature would allow advertisers to specify contexts in which they would prefer their messages not appear.

(What those contexts might be — political discussions? medical queries? conversations about competitors? — the company declined to say.)

The development comes as advertisers who have begun buying space within ChatGPT's conversational interface have expressed frustration with what several described as limited controls over placement and scant visibility into where their ads actually ran. Four executives involved in buying ads on the platform, speaking with Adweek, cited the difficulty of explaining their target audiences to a system that was, after all, designed to understand natural language.

OpenAI, which has moved rapidly from research laboratory to consumer platform to advertising venue in the span of roughly two years, declined to offer a timeline for when the negative targeting feature might become broadly available. The product remains in development.

For the advertising industry, the addition of exclusion tools to ChatGPT would represent a familiar step in the maturation of a new medium. Television, the web, social media and programmatic display all eventually developed elaborate systems for helping advertisers avoid appearing next to content they found objectionable — or simply off-strategy.

That OpenAI is following the same path suggests that even the most novel of platforms must eventually answer the oldest of advertiser questions: not just who will see this ad, but what will they be thinking about when they do?

Original story published in adweek.com: "EXCLUSIVE: OpenAI Is Testing Exclusion Targeting for ChatGPT Ads"