ChatGPT Ads Now Rewrite Your Ads by Default. Here's the Switch.
Updated October 30, 2026
Starting October 31, 2026, ChatGPT ads have a setting called creative enhancements turned on by default. It can rewrite your headlines and descriptions, translate them, resize images and send clicks to a different page on your site, and those versions can run without your approval. If your practice runs ChatGPT ads, you can turn it off in Ads Manager today.
Here's why that matters for a dental or speech-language pathology practice. The ad still carries your practice's name and your license, and dental advertising rules are written about what the ad says.
Do this today
Sign in to OpenAI Ads Manager (the dashboard where ChatGPT ads are set up), open Settings and switch off the Creative enhancements card. OpenAI's help article shows the setting.
Open each campaign and check its own setting. A campaign setting overrides the account default.
Not sure who runs your ads or where? Ask today, and ask about Google and Microsoft too. Both have their own version of this, covered below.
What are ChatGPT ads creative enhancements?
Creative enhancements is OpenAI's name for a setting that lets ChatGPT adjust your ad to fit each conversation. It works from two things you already gave it: your ad and your landing page (the page someone sees after they click).
On October 20, OpenAI emailed advertisers that it would turn the setting on for new and existing campaigns. Its notices have given start dates from October 29 to October 31, 2026, so assume it's on unless someone switched it off.
When it's on, the system can make four kinds of changes:
Headlines and descriptions: new or rewritten text meant to connect your ad to the conversation.
Translation: your ad text in another language.
Images: resized to fit the ad format.
Destination URL: the click can go to a different page on your website than the one you picked.
OpenAI says these versions can run without you reviewing or approving each one, and Ads Manager doesn't show a preview of every variation. Back in July, ChatGPT ads could already draft ads from a website, but each one needed the advertiser's approval before it ran. This setting keeps the drafting and drops that step.
How do I turn off creative enhancements in ChatGPT ads?
Sign in to OpenAI Ads Manager and open Settings.
Find the Creative enhancements card and switch it off. This sets the default for the whole account.
Open each campaign and check its setting. OpenAI says campaign settings take priority over the account default, so one campaign can still be on.
If you opted out before, OpenAI says existing account and campaign opt-outs are kept. Check anyway.
Save a screenshot with the date, so you know when it was turned off and by whom.
There's one switch for all four changes, so you can't keep the image resizing and turn off the rewriting. Reporting also doesn't show which version ran in a given impression. I'd leave it off on any ad that names a procedure, a doctor or a result.
Why does this matter for a dental practice?
A shoe brand can let software try new headlines. A practice ad is a professional ad, and the rules look at the words. Texas's dental board rule, for example, says every ad is presumed to have been approved by the licensee named in it.
Ask your attorney how your state treats an ad a platform rewrote. Here's where I'd look first:
What the AI can change | What could go wrong | What to check |
|---|---|---|
Headline and description | A new line says "specialist," "best" or "pain-free," or predicts a result. | Your ads and landing pages for the words to watch below. The AI writes from them. |
Translation | Nobody on your team can read the new version, and you can't preview every one. | Turning the setting off is the only way to stop translation. |
Image resizing | A crop cuts off the doctor's name or a disclaimer printed in the image. | Keep required wording in the ad text and on the landing page, not only in the image. |
Destination URL | A click lands on a page you didn't pick, like an expired offer or a before-and-after gallery. | Every page on your site is one you'd be fine sending a new patient to. |
Which words should a dental or SLP practice watch for?
These are the lines I'd check first in any ad, written by a person or by AI. The examples come from Florida, Texas and the American Dental Association (ADA). Your state's dental board has its own rules.
"Specialist." Florida law lets a dentist hold out as a specialist only with ADA-recognized specialty training or board status (Fla. Stat. 466.0282). The ADA recognizes 12 specialties, including oral and maxillofacial surgery and orthodontics, and its Code says general dentists announcing services should state that they are general dentists.
"Board certified." In Texas, only for certifying boards the dental board's rule recognizes (22 TAC 108.54).
Results. "Perfect smile," "pain-free" or any prediction. Florida bars ads likely to create "false, unjustified expectations of favorable results" (Fla. Stat. 466.019).
Comparisons and praise. "Best," "top" or "better than." Florida's same statute bars laudatory statements and quality comparisons, and the ADA Code treats unsupported superiority claims as misleading.
Reviews and before-and-after photos. Texas allows testimonials only from patients of record, and photos of actual patients only with written consent and the line "Actual results may vary" (22 TAC 108.57).
For speech-language pathologists, ASHA's Code of Ethics bars misrepresentation in advertising and says members shall not guarantee the results of any treatment. Florida licenses SLPs under a different chapter than dentists, so check that board's rules too.
For the full Florida list with safer rewrites, see my Florida dental ad claims checklist.
Can dental practices run ChatGPT ads at all?
Only with OpenAI's approval. Its ad policy says ads for clinical care providers are currently disallowed, and that it may approve healthcare advertisers case by case through manual review.
Search Engine Journal reported in August that dental services were among the health categories OpenAI listed. The policy page I read on October 30 didn't name dental, so expect that review if you apply. ChatGPT also keeps ads out of mental health and personal health conversations.
For how ChatGPT answers fit with Google's AI results, see AI search for practices.
Do Google Ads and Microsoft Ads do this too?
Yes, in their own ways. Both apply to responsive search ads, the text ads built from a set of headlines and descriptions you write.
Google calls it text customization, part of AI Max for Search. It used to be called automatically created assets. Google says customized text can serve as soon as it's generated, and you review it afterward in reports (Google's text customization FAQ).
Microsoft Advertising started adding autogenerated assets to responsive search ads that don't have the full set of headlines and descriptions. Microsoft says sensitive industries are excluded and you can opt out in campaign settings (Microsoft's April 2026 update).
What should we check this week?
Turn off ChatGPT creative enhancements for the account, then check every campaign.
Check Google text customization and Microsoft autogenerated assets on every campaign.
Read your landing pages like ad copy, since the AI writes from them. Take down expired offers and any line you wouldn't run in an ad.
Make sure every page a click could reach is current, including old offer pages and galleries.
Name one person who approves ad copy before it runs, with the doctor signing off on anything clinical.
Keep dated copies or screenshots of each ad version you approve.
Have your attorney check the words above against your state's rules.
Want a second look at where your ads run?
I've been in marketing since 2006, and I'm a marketer, not a lawyer. What I can do is help you see where your ads run, what they say and where each click lands.
Start with a free 15-minute practice review. You keep your written Double Jaw Method Plan either way.
Book your free practice review
This is general information, not legal advice. Rules differ by state and change over time, so have your attorney review your ads before you rely on it.
Common questions
What are creative enhancements in ChatGPT ads?
A setting that lets ChatGPT rewrite your ad's headlines and descriptions, translate the text, resize images and change which page on your site a click goes to. OpenAI turns it on by default starting October 31, 2026, and the versions can run without your approval.
How do I turn off creative enhancements?
In OpenAI Ads Manager, open Settings and switch off the Creative enhancements card. Then open each campaign, because a campaign setting overrides the account default.
Can a dental practice advertise on ChatGPT?
Only with approval. OpenAI's ad policy says ads for clinical care providers are currently disallowed, with healthcare advertisers reviewed case by case. Ads also stay out of mental health and personal health conversations.
Do Google Ads and Microsoft Ads rewrite ads too?
Both can add AI-written text to responsive search ads. Google calls it text customization in AI Max, and Microsoft calls it autogenerated assets. Check the setting on every campaign.
Can a general dentist's ad say "specialist"?
In Florida, only dentists with ADA-recognized specialty training or board status may hold out as specialists, and the ADA Code asks general dentists to say they are general dentists. Other states have their own rules, so ask your attorney.
