OpenAI has removed manual bid inputs from ChatGPT Ads entirely for advertisers who opt into a new strategy called Maximize results, letting the system set and adjust bids on its own toward a chosen campaign goal. The same update lets advertisers restrict a campaign to specific surfaces: the iOS app, the Android app or the web version of ChatGPT. Together the two changes mark a shift from an advertiser steering individual bids to an advertiser steering only the goal, while the system decides how and where to spend toward it.
That tradeoff is not new to digital advertising. Google made a similar bet with Performance Max and Smart Bidding years ago, trading advertiser control for scale on the promise that the algorithm sees more signal than a human ever could. What is different here is the surface. A conversational answer has none of the fixed layout of a results page, so an advertiser handing bid control to ChatGPT Ads is also handing over judgment about placement in an environment where there is no obvious slot to audit.
The Maximize results option is designed to wring the largest possible number of outcomes out of a fixed budget, according to Search Engine Land. An advertiser sets the goal; the bidding decisions that used to sit with a media buyer now sit with the platform.
Platform-level targeting solves a narrower, more practical problem. An advertiser who sees stronger conversion rates or higher customer value on mobile can now confine delivery to the iOS or Android app rather than spreading budget across a web experience that performs differently. That control did not previously exist inside ChatGPT Ads campaign setup.
Two other reporting changes round out the release. Advertisers can now connect campaign data to WorkMagic to compare ChatGPT Ads performance against other channels, and that integration can pass conversion outcomes back to OpenAI through the Conversions API, giving the bidding system another feed of results to optimize against.
Both additions build on a run of updates this publication has already tracked. ChatGPT Ads picked up oCPC conversion bidding and product carousels on August 10, and a one-day view-through conversion metric followed on August 20, letting advertisers see conversions credited to an ad exposure without a click. The same August 10 report said Brazil and Mexico access was expected within days; this release confirms that expansion has now landed, roughly two weeks later than first indicated.
None of the reporting so far includes independent measurement of how Maximize results performs against manually managed bids, or how much of the platform’s conversion count now comes from view-through credit rather than clicks. Those two gaps sit next to each other for a reason: an automated bid strategy optimizing toward a conversion definition that now includes view-through credit could report strong results without a proportional increase in incremental demand.
Search teams running budget on ChatGPT Ads should test Maximize results against a manually managed control group before reallocating spend, and should isolate view-through conversions in any performance comparison. The platform-level targeting split is the more immediately useful change for teams that already suspect their app and web audiences convert differently, since it requires no trust in a new automated system to act on.
Search Engine Land reported these ChatGPT Ads updates on August 24, 2026, in a story by Anu Adegbola.
