OpenAI will switch on creative enhancements for every ChatGPT Ads campaign unless you opt out

What changed

Week 43 follow-up: 2026-10-20

Earlier coverage and context (2026-10-20)

ChatGPT Ads creative enhancements can adapt existing advertisements using their copy and landing-page content. OpenAI's help page describes changes to headlines, descriptions, image sizing, language and destination URLs. Enabled variations can serve without approval of every version, making this an account-setting decision for advertisers with controlled messaging.

There is a material date discrepancy. The official help page checked on 21 October says automatic activation starts on 29 October 2026. Barry Schwartz's first-person report, published on 20 October, reproduces an advertiser email giving 30 October. Thomas Eccel also shared an account observation of the setting. These sources support the feature and the need to inspect settings, but do not establish one universal activation date.

OpenAI says existing account and campaign opt-outs are preserved, with campaign-specific settings taking priority over the account default. Advertisers can change the setting in Ads Manager. The documentation does not provide a preview of every generated variation. The accompanying account-setting screenshot is credited to Barry Schwartz, separately from Eccel's account observation. First-ever reporting priority has not been established.

PMC's practical recommendation is to review the setting before the earlier documented date and retain the notice received by the specific account. This is a precaution against the conflicting timetable, not a claim that both dates apply to every advertiser. Teams using legally reviewed offers or carefully qualified claims should assign an owner to that review.

Audit the source material as well as the switch: expired promotions, inconsistent prices and unclear product descriptions create avoidable uncertainty in any automated creative workflow. Record account and campaign choices together, then compare subsequent delivery with a dated baseline. A change in overall results alone would not isolate the effect of creative enhancements. The immediate task is to make an informed setting choice and keep its evidence, rather than assume a visible toggle proves every variation has already served.