Menu Close

OpenAI to test visual ads in ChatGPT image generation

News illustration: on a plum-to-apricot gradient, a white ChatGPT image-generation card with the OpenAI logo shows a generated sunset landscape above a prompt bar; across a dashed gap sits a separate white tile labeled Sponsored with a generic orange sneaker and a green verification check; chips VISUAL ADS and US TEST.

OpenAI will begin testing image-based ads in ChatGPT later this month, starting with the moments when people generate pictures. The new format, announced Monday in an OpenAI post, will show images of sponsored products and services, and U.S. testing starts with an initial group of advertisers.

The ads will be clearly labeled and kept separate from the image ChatGPT is creating, OpenAI said, and advertising does not influence the answers ChatGPT gives. Until now, ChatGPT ads, which arrived in February, showed only a company’s name, logo and link in a sponsored section beneath the chat, The Verge reported. The Verge said the ads will not appear for subscribers to ChatGPT’s Plus, Pro or Enterprise plans.

For advertisers, OpenAI is adding outside checks. It is starting brand-suitability evaluation pilots with ad-verification firms DoubleVerify and Integral Ad Science, which will assess how its brand-safety rules are applied in a controlled testing environment. OpenAI said the partners will not have access to private user conversations. “ChatGPT is creating a fundamentally different advertising environment, where context is defined by the conversation itself rather than a page or video,” DoubleVerify chief executive Mark Zagorski said in the post.

OpenAI also expanded its measurement tools. Advertisers can now send conversion data from Hightouch, Tealium and LiveRamp, and track results through attribution firms including AppsFlyer, Adjust, Triple Whale, DV Rockerbox and Northbeam. OpenAI is running early geo-based incrementality experiments with Haus, Measured and WorkMagic. It cited partner figures showing WeightWatchers’ cost per acquisition on ChatGPT ads was 15.3% below its blended paid-search benchmark, according to DV Rockerbox.

Placement guardrails are meant to keep ads out of emotionally vulnerable, sensitive or otherwise unsuitable conversations, and qualifying advertisers can now block their ads from appearing near specific phrases. OpenAI says ChatGPT reaches 1.2 billion people each week.

The visual format extends an ad business that OpenAI said last month had reached a $1 billion annualized run rate.

Sources

0 0 votes
Article Rating
Subscribe
Notify of
0 Comments
Inline Feedbacks
View all comments
0
Would love your thoughts, please comment.x
()
x