What Is OpenAI Changing?
OpenAI is making automatic ad variations the default in ChatGPT Ads, its advertising product that places ads below ChatGPT’s answers. The feature, called creative enhancements, is already a setting in Ads Manager, and OpenAI’s email refers to advertisers’ “existing account and campaign-level opt-outs.” What changes at the end of October is that it will be on unless an advertiser has turned it off.
OpenAI describes the purpose as making ads “more relevant to the conversation and user interests, helping to improve performance while reducing work to create multiple versions.” It says: “We use your existing ad and landing-page content to adapt your message” (OpenAI Help Center). The system can take into account three inputs, in OpenAI’s list: “The context of the conversation or user interests,” “Your existing ad,” and “Content from the ad’s landing page.”
The variations need not be written ahead of time. “Variations may be prepared in advance or created in real time based on the conversation. When enabled, these ads can run automatically without requiring you to review or approve each version” (OpenAI Help Center).
ChatGPT Ads is still in beta. OpenAI says it ran “a focused pilot from February through April” and is now “expanding access through self-serve onboarding in Ads Manager Beta,” and that during the beta it will keep changing “delivery systems, inventory, ad formats, and how advertisers buy, manage, measure, and optimize campaigns” (OpenAI, ChatGPT Ads FAQ).
Which Date Is Right: October 29 or October 30?
OpenAI has published both, so advertisers should treat October 29 as the deadline. The help page reads: “Starting Thursday, October 29, 2026, creative enhancements will be enabled automatically for new and existing campaigns. You can opt out before October 29 or change your settings at any time in Ads Manager” (OpenAI Help Center).
The email OpenAI sent to advertisers says something slightly different. Barry Schwartz, executive editor of Search Engine Roundtable, received it and published its text and a screenshot on October 20, 2026: “Starting Friday, October 30, 2026, we’ll enable creative enhancements for your new and existing self-serve campaigns” (Search Engine Roundtable). PPC Land repeated the October 30 date on October 21, citing that report (PPC Land).
| OpenAI source | Start date | Which campaigns | Opt out before |
|---|---|---|---|
| Help Center article, read October 21, 2026 | Thursday, October 29, 2026 | “new and existing campaigns” | October 29 |
| Email to advertisers, as published October 20, 2026 | Friday, October 30, 2026 | “new and existing self-serve campaigns” | October 30 |
The two sources also differ on scope: the email names self-serve campaigns, while the help page does not qualify “campaigns.” OpenAI has not explained either difference publicly, and we could not find a correction. Our recommendation, not an OpenAI rule: set the opt-out you want by October 28, and check the setting in your own account rather than relying on either date.
What Can the Enhancements Change?
Four elements, according to OpenAI’s list. The table quotes the help page; the right-hand column is our reading of what each change means for an advertiser.
| Element | What OpenAI says may happen | What to watch |
|---|---|---|
| Headlines and descriptions | “We may create new or updated text to connect your offering to the conversation or user interests.” | Wording you did not write, built partly from your landing page |
| Language | “We may translate your ad text into another language.” | Claims and product names in languages your team may not review |
| Images | “Images may be resized to fit the ad format.” | Cropping that cuts logos, text or required disclosures from an image |
| Destination URLs | “The destination URL may be optimized to direct people to a relevant page on your website.” | Clicks landing on pages you did not choose, which may not carry your tracking |
Source: OpenAI Help Center, “Creative enhancements for ChatGPT Ads,” read October 21, 2026. OpenAI’s example turns the headline “Fresh groceries delivered” into “Groceries for quick weeknight dinners” for a conversation about weeknight meals, and notes that the example “assumes the original ad and landing page support the product and delivery details shown” (OpenAI Help Center).
This is separate from a feature OpenAI calls suggested ad drafts, which prefills a draft from website metadata for the advertiser to review. OpenAI says that tool “does not generate new copy or imagery with AI” (OpenAI, Create Ads for ChatGPT Ads). Creative enhancements do.
How Do You Opt Out?
In Ads Manager, at the account level or campaign by campaign. OpenAI says advertisers can opt out before the start date “or change your settings at any time in Ads Manager,” and that “Existing account- and campaign-level opt-outs are preserved. Campaign-specific settings take priority over your account default” (OpenAI Help Center).
Search Engine Roundtable published a screenshot of the setting in an ad account’s settings section, already shown as toggled on for that account (Search Engine Roundtable). Because campaign settings override the account default, an advertiser who wants one strictly controlled campaign and automatic variations elsewhere can set the account to on and that campaign to off, or the reverse.
Who Is Responsible for What a Rewritten Ad Says?
OpenAI’s rules still apply to every version that runs. Its Ads Policies say: “Ads must be truthful and not misleading, including unfounded claims about capabilities, pricing, outcomes, affiliations, or comparisons with other products or companies” (OpenAI Ads Policies). The help page tells advertisers to “Keep your ads and landing pages current, including prices, promotions, product details, and availability,” and to make sure “your creative and landing pages comply with OpenAI’s Ads Policies” (OpenAI Help Center).
The practical consequence, in our reading: the landing page becomes source copy. An expired promotion, an old price, a customer quote or a superlative that sits on the page could be lifted into a headline the advertiser never saw. The help page does not say who is liable for a rewritten ad that breaks a rule, and we have not reviewed OpenAI’s advertising terms on that point. Advertisers who make regulated or comparative claims should get their own advice on that question; for the general US rules on paid content, see what the FTC requires when a placement is paid for. This is not legal advice.
Some advertisers reacted badly. Melissa Mackey, a paid-search practitioner, wrote on X: “We ended up pausing our ChatGPT ads test due to poor lead quality and now I’m glad we did. Not cool” (quoted by Search Engine Roundtable). That is one advertiser’s view, not evidence about performance.
Does This Change What ChatGPT Says About a Company?
No, according to OpenAI. The help page’s FAQ asks whether creative enhancements change ChatGPT’s answers and replies: “No. Ads do not influence ChatGPT’s answers. Ads are separate and clearly labeled” (OpenAI Help Center). OpenAI’s overview says ads appear below ChatGPT responses and are not shown to users on Plus, Pro or any Business plan, or to accounts identified as belonging to users under 18 (OpenAI, Ads in ChatGPT: The Basics).
That separation matters for anyone weighing paid against earned visibility in AI answers. Buying a ChatGPT ad buys a labeled slot under an answer; it does not buy a mention inside the answer, and an automatically rewritten ad does not change that. Our earlier report on how many businesses advertise on ChatGPT covers the same line.
Can You Measure Whether It Works?
Only roughly. OpenAI says Ads Manager reporting covers “impressions, clicks, spend, CTR, and attributed conversions if conversion measurement is set up” at campaign, ad group and ad level, and warns: “Changes over time do not, on their own, establish the effect of creative enhancements” (OpenAI Help Center). The help page does not say whether reports will show which variation ran or what it said.
Tracking needs a check too. OpenAI says static tracking parameters such as UTM tags added to landing-page URLs “persist on ad clicks” (OpenAI, Ads in ChatGPT: The Basics). Its pages do not say whether an “optimized” destination URL keeps the parameters from the original. Until OpenAI says so, compare ChatGPT Ads clicks with your analytics after October 29 and look for landing pages you did not choose; see how to track links with UTM parameters.
What This Means for Anyone Buying a Placement
These are our recommendations, not OpenAI rules.
- Decide before October 29. If legal, medical, financial or regulated claims need sign-off in your business, turn creative enhancements off at the account level and turn them on only where variation is acceptable.
- Audit the landing pages. Remove expired offers, unsupported superlatives and out-of-date prices from any page an ad points to, and from pages nearby on the site that a “relevant page” could mean.
- Check images for cropped text. If an image carries a disclosure or a required logo, test how it looks resized, or use images without embedded text.
- Keep paid and earned separate in reporting. A ChatGPT ad is advertising, labeled as such. Do not count ad impressions as AI visibility for your brand, and do not let anyone sell it to you that way.
Methodology and Limitations
Four limits apply. First, OpenAI’s Help Center refused direct requests from our tools (HTTP 403), so we read the creative-enhancements page and three related OpenAI articles through a text-extraction reader on October 21, 2026; the wording quoted matches the passages Search Engine Roundtable reproduced. Second, we did not see OpenAI’s email ourselves; we rely on the text and screenshot Search Engine Roundtable published. Third, we have no ChatGPT Ads account and did not test the setting. Fourth, OpenAI says the beta will keep changing, so dates and features may move.
Conclusion
From the end of October, a ChatGPT ad can say things its advertiser never wrote, in languages it never chose, pointing to pages it never picked, unless the advertiser opts out. That may well lift performance, as OpenAI argues, but OpenAI has published no data showing it does, and the rules on truthful claims still apply to every variation. The cautious course is to opt out by October 28, fix the landing pages, and switch the feature back on campaign by campaign once you can see what it writes.
Frequently Asked Questions
What are creative enhancements in ChatGPT Ads?
They are automatic variations of an ad. OpenAI says they can include new or updated headlines and descriptions, image resizing, destination-URL optimization and translation, using the existing ad, its landing page and the conversation.
When does OpenAI turn on creative enhancements by default?
OpenAI’s help page says Thursday, October 29, 2026; its email to advertisers, as published by Search Engine Roundtable, says Friday, October 30, 2026. Treat October 29 as the deadline.
How do I opt out of creative enhancements?
Turn the setting off in Ads Manager, at the account level or for individual campaigns. OpenAI says existing opt-outs are preserved and campaign settings take priority over the account default.
Can I preview each version of my ad?
No. OpenAI says advertisers do not need to review or approve each enhanced version and that the experience “does not provide a preview of every variation.”
Do ChatGPT ads change ChatGPT’s answers?
OpenAI says no: “Ads do not influence ChatGPT’s answers. Ads are separate and clearly labeled.” Ads appear below responses.
Who is responsible if a rewritten ad makes a misleading claim?
OpenAI’s Ads Policies require ads to be truthful and not misleading, and its help page tells advertisers to keep landing pages accurate. Its help page does not address liability for rewritten ads. This is not legal advice.
Sources and Methodology
This guide is built from four OpenAI Help Center articles and OpenAI’s Ads Policies, plus two trade reports, all read on October 21, 2026. The checklist and the recommendation to treat October 29 as the deadline are ours, not OpenAI’s. We did not test the feature.
- platform documentation OpenAI Help Center, Creative enhancements for ChatGPT Ads. Supports the October 29 date, inputs, what can change, opt-out rules, no-preview and measurement statements. Read October 21, 2026 (via a text reader; direct requests returned HTTP 403).
- platform documentation OpenAI Help Center, Ads in ChatGPT: The Basics. Supports ad placement below responses, excluded users and UTM persistence. Read October 21, 2026.
- platform documentation OpenAI Help Center, Create Ads for ChatGPT Ads. Supports the distinction from suggested ad drafts. Read October 21, 2026.
- platform documentation OpenAI Help Center, Frequently asked questions (ChatGPT Ads). Supports beta status and the February to April pilot. Read October 21, 2026.
- platform documentation OpenAI, Ads Policies. Supports the truthful-claims rule. Read October 21, 2026.
- trade press Barry Schwartz, Search Engine Roundtable, OpenAI Auto Enabling Creative Enhancements In ChatGPT Ads, October 20, 2026. Supports the email text, October 30 date, settings screenshot and the advertiser quote. Read October 21, 2026.
- trade press Luis Rijo, PPC Land, Netflix cuts its ad target to $2.7bn as OpenAI turns on automatic ad rewrites, October 21, 2026. Supports second report of the October 30 date. Read October 21, 2026.