ChatGPT Ads Will Auto-Rewrite Your Copy: How to Opt Out

If you advertise on ChatGPT, or plan to, there is a setting you need to look at this week. OpenAI is turning on creative enhancements for ChatGPT ads by default, and the system can change your ads without you reviewing each version.

What OpenAI Is Switching On

According to PPC Land's report on OpenAI's October 20 email, the enhancements can rewrite headlines and descriptions, translate ad text, resize images and change destination URLs. The Creative enhancement toggle in Ads Manager is set to On by default. The email says the change starts Friday, October 30. OpenAI's help page, updated shortly before PPC Land checked it, gives Thursday, October 29. Since the two sources conflict, plan around the earlier date.

You can opt out at the account level or the campaign level, and campaign-level settings take priority. Existing opt-outs carry over. The email limits the change to self-serve campaigns, but the help page does not, and it is unclear whether campaigns bought through sales teams or technology partners are included. If you buy through a partner, ask them directly.

Why the Destination URL Is the Part to Worry About

Rewritten headlines are a brand-voice risk. A changed destination URL is a measurement and compliance risk. If a system can swap the page an ad points to, your UTM parameters, conversion tracking and landing page promise can all stop lining up with the ad copy someone approved. For local and service businesses, rewritten copy can also drift into claims about pricing, licensing, availability or results that you never signed off on.

What Else Changed This Month

The default is arriving alongside a fast expansion. OpenAI says ChatGPT ads are now sold in 89 countries after adding 26, and that more than 100,000 advertisers have run campaigns since ads launched just over eight months ago. That is a cumulative figure, not an active-advertiser count, and OpenAI did not publish spend or pricing. Ads appear only on the Free and Go plans.

OpenAI also announced visual display ads that appear alongside images users ask ChatGPT to generate. According to TechCrunch, they begin appearing later in October in the U.S. for an initial group of advertisers, and OpenAI says they will be clearly labeled and will not influence ChatGPT's answers. Measurement partners named include AppsFlyer, Triple Whale, Adjust, Northbeam, Singular and Kochava for click attribution, and Haus, Measured and WorkMagic for geo experiments. DoubleVerify and Integral Ad Science are involved in brand suitability pilots.

What We Would Do

This is our recommendation, not OpenAI's. Opt out at the account level before the earlier deadline. Then, if you want to test the feature, turn it on for one campaign, review every generated variation in Ads Manager, and compare results against an identical campaign with enhancements off. Confirm your tracking parameters survive on the live ads. Keep the test small, and set the success threshold against your own break-even return, not a platform benchmark.

Frequently Asked Questions

Will opting out hurt delivery or cost?

OpenAI has not said. Test it yourself instead of assuming either way.

Does this affect agency-managed or partner-bought campaigns?

It is unclear. The email and the help page disagree on scope, so ask your agency or ad partner to confirm in writing.

Is ChatGPT advertising worth testing for a small business?

Possibly, as a small, bounded test. We walked through how to size one in our post on whether to advertise on ChatGPT.

If you want help setting up a controlled test, that is what our paid media team does. The same default-on pattern is showing up at Meta, which we covered in our look at Meta's new AI ad terms.

Sources: PPC Land on default rewrites, PPC Land on 89 countries, TechCrunch on visual ads.