ChatGPT Ads now rewrites your ads by default. What global brands should check

ChatGPT Ads is growing fast. On 21 October 2026, OpenAI expanded ChatGPT ads to 26 more countries, bringing the total to 89, and said more than 100,000 advertisers have run campaigns since launch. Nine days later, on 30 October, a feature called creative enhancements was switched on by default for campaigns that had not opted out. It lets ChatGPT rewrite your headlines and descriptions, translate your ads, resize your images and change where your ad links to.

Put those two changes together and more brands, in more markets and languages, now have ads running in versions they did not write and did not approve. This article explains what creative enhancements can change, where the risks are for global brands, and what to check in your account.

What changed

What creative enhancements can change

Creative enhancements use the user's conversation or interests, plus your ad and landing page content, to adapt your ad. Versions can be made in advance or in real time.

Three details make this harder to manage than it sounds:

• No review step. According to OpenAI, enhanced ads can run without you reviewing or approving each version, and there is no preview of every variation.

• One switch for everything. The setting covers all four types of change. You cannot, for example, allow image resizing but block rewritten text. You can set it at account level and override it per campaign.

• No variant reporting. Ads Manager does not show which rewritten headline, language or landing page ran in a given impression, so you cannot see which version drove the result. Turning the feature off stops new changes but does not undo versions that already ran.

Why it matters more now that ChatGPT runs in 89 countries

For a brand running ads in one market and one language, a rewritten headline is a small risk. For a global brand, it multiplies.

• Legal text differs per market. A financial, pharma, alcohol or energy ad often needs exact wording that changes by country. A rewrite or translation can drop or change it, and the brand is still responsible for what runs.

• Translation skips local review. Many brands have local teams check every translated ad. Automatic translation goes around that step.

• Landing pages may not match. If ChatGPT sends a Colombian user to a page built for the US, the price, currency or offer may be wrong.

• Brand consistency across channels. Your ChatGPT ad may now say something different from your approved social, display and CTV ads for the same campaign.

This is part of a wider pattern. Ad platforms such as Meta, Google and TikTok increasingly use their own AI to adapt creative at the point of delivery. Each platform's AI works inside its own walls. The brand is the only one that sees all of them, and the brand carries the risk. We covered the quality side of this in Why AI-Generated Ads Have a Quality Problem (And How to Fix It).

What to check in your ChatGPT Ads account

1. Check the account default. Go to Manage settings in Ads Manager and see whether creative enhancements are on. Decide on purpose instead of keeping the default.

2. Turn it off per campaign where it matters. Regulated products, campaigns with legal disclaimers, price-led offers and markets without a localised site are the first candidates.

3. Put UTM parameters on the ad URL. Parameters set on the ad URL are not overwritten, and they help you spot when clicks land on a different page.

4. Keep your landing pages current. If ChatGPT can send clicks to any relevant page on your site, every page needs up-to-date prices and offers.

5. Watch results by market. Without variant reporting, a sudden change in click-through or conversion rate in one country may be your only signal that something changed.

6. Write the decision down. Record which campaigns allow enhancements and why, so legal and brand teams know what can run.

Where Cape.io fits

You cannot control what each platform's AI does after your ad leaves your hands. You can control what goes in. The closer your source creative is to right, in every market and language, the less any platform has to fix.

• Approved versions per market. Cape.io generates localised versions from one master template, with the right language, currency and legal text for each market, so you do not depend on platform translation.

• Checked before it goes live. Every asset is checked against brand rules, platform specs and legal requirements, including rules from the ASA and the EU AI Act, before it is delivered. See What the EU AI Act Means for Your Ad Creative Workflow.

• One source for every channel. The same approved creative goes to social, display, CTV and broadcast, so your message stays consistent wherever a platform adapts it.

Want to see how your team can keep control of creative across 89 markets? Book a demo with Cape.io.