Ask ChatGPT for a picture later this month and you might see an ad next to it. OpenAI says it will test a new visual ad format during image generation in the US, starting with a small group of advertisers. If you use the free or $8 Go plan, ChatGPT is about to fit ads into a creative task, where until now they've appeared below its text answers.
OpenAI says the ads will be clearly labeled and kept separate from the images ChatGPT makes for you. Brands will use them to show off products, how people use them, or the experiences they're selling. The company repeated its central promise too: "advertising does not influence the answers ChatGPT provides".
The real audience for this update is advertisers
The new format is the part you'll see. Most of the announcement is about something else: showing advertisers that ChatGPT ads pay off. OpenAI is hooking up data tools such as LiveRamp, Tealium and Hightouch and attribution firms like AppsFlyer and Branch, so brands can track conversions (sales or sign-ups that follow an ad). It's also testing with Haus and Measured to find out whether ads cause purchases or just get credit for ones that would have happened anyway. The early results come from partners and OpenAI. WeightWatchers, for example, paid 15.3% less per new customer than its usual paid search benchmark, according to measurement firm DV Rockerbox in figures OpenAI published with the announcement. ContentGrip notes the figure is preliminary.
There are new safety checks as well, and they're built for brands. DoubleVerify and Integral Ad Science will review how OpenAI applies its brand-safety rules "without accessing private user conversations". Advertisers can also add "negative phrases" so their ads stay out of chats that mention words they want to avoid. OpenAI describes those outside reviews as brand safety checks, and its announcement doesn't name anyone outside the company testing whether ads stay out of the answers themselves.
What happens to your chats, and how to opt out
OpenAI's rules haven't changed since ads launched in February. By default, it picks ads using the topic you're discussing, your past chats and how you've reacted to ads before. Advertisers get totals, such as how many people viewed or clicked, and never see your chat history. You can turn off personalization, delete your ad data and hide individual ads. Ads also shouldn't appear in chats about health, mental health or politics, or for users OpenAI thinks are under 18.
To get rid of ads completely, you can upgrade to Plus or a higher plan. On the free plan, you can switch to an ad-free mode that gives you fewer messages per day.
The test starts later in October with a small group of US advertisers. The next thing to watch is whether OpenAI widens it to more brands or other countries once those first results come in, and whether the measurement numbers hold up outside early partner reports.
Sources
- 1.OpenAI Adds Visual Format, Makes ChatGPT Ads More Measurable · MediaPost
- 2.ChatGPT Ads adds visual format and measurement · ContentGrip
- 3.ChatGPT Now Has Ads for Free and Go Tier Users · MacRumors
- 4.You can avoid ChatGPT ads, but your Free limits may change · Yahoo Tech
- 5.Building advertising for the way people use AI · OpenAI
- 6.Introducing ChatGPT Go, now available worldwide · OpenAI
- 7.Ads in ChatGPT · OpenAI Help Center
Reported by the WattsUpNext desk from the sources linked below. Spot an error? Tell us at corrections@wattsupnext.com.
The WattsUpNext Brief
Get stories like this in your inbox.
One email each weekday, only the topics you choose. Real news, sourced, no fluff. Unsubscribe in one click.
