Google last month rolled out a new branded search control inside AI Max for Search campaigns, giving advertisers a native toggle to decide whether their ads appear on queries containing brand names. The feature, first spotted and shared on LinkedIn by Thomas Eccel, an ads specialist, surfaces three distinct options within the AI Max campaign interface - a level of granularity that had been absent from the format since it launched in May 2025.

The addition is modest in scope but consequential in practice. AI Max has drawn persistent criticism from practitioners for expanding into queries advertisers would rather avoid, including branded terms belonging to competitors and, in some cases, the advertiser's own brand. The new control addresses that directly.

What the three options do

According to Google, the branded searches setting reads: "Choose how you want your ads to show on searches that contain brand names." The three available options are: show ads on all relevant searches, control branded searches with brand inclusions and exclusions, and show ads only on unbranded searches.

The first option reflects the default behaviour that AI Max has operated under since launch. The second gives advertisers a more surgical path, preserving the ability to include some branded terms while excluding others - for instance, allowing the advertiser's own brand while blocking competitors, or the reverse.

The third option is the most absolute. According to Google's own interface text, advertisers should "use when you only want ads to show on generic searches, avoiding all brand names known to Google." The qualification "known to Google" is notable: it signals that the system draws on its own index of brand entities rather than relying on a manually maintained list supplied by the advertiser. Ads running under this setting will not appear on queries containing the advertiser's brand, competitor brands, or any other brand the system recognises.

Google also flags in the interface that the unbranded-only setting "is highly restrictive." Its note adds that if the goal is only to prevent ads from showing on the advertiser's own brand - rather than all brands - a brand exclusion is the more appropriate mechanism.

Why separate branded campaigns break down under AI Max

The structural problem this feature is designed to solve is well understood among paid search practitioners. Many advertisers maintain parallel campaigns: one targeting branded queries, and one targeting non-branded or generic terms. The branded campaign typically carries a different target ROAS, tightly controlled impression share, and isolated conversion data used for attribution modelling.

When AI Max is active in a generic or prospecting campaign, it expands targeting based on inferred intent rather than exact keyword matching. That expansion has been documented to pull in branded queries - both the advertiser's own brand and competitors' - at CPCs higher than those a dedicated branded campaign would ordinarily deliver. The result is cannibalization of the branded campaign's traffic and muddied attribution data.

Thomas Eccel described the problem plainly on LinkedIn: "If you're running a separate branded Search campaign (which you should be) you 100% don't want AI Max cannibalizing those clicks at a higher CPC and breaking your attribution."

He added that the unbranded setting "lets you keep AI Max purely in prospecting mode," and contrasted it with the previous workaround: "Before this, your only protection was a brand exclusion list. Now it's a native setting inside the campaign."

The reaction from practitioners in the LinkedIn thread was broadly positive. Ashis Patra, founder of Paidbox, wrote that the feature is "a meaningful update for advertisers running separate brand and non-brand strategies," and that "having native control over branded queries should provide cleaner attribution and better visibility into incremental growth."

Ryan Parkes, a paid media professional, flagged a practical edge case: a campaign targeting searches like "Honda Civic" raises the question of whether "Honda" alone would be classified as a brand exclusion, potentially preventing the ad from appearing on a legitimate product search. The concern reflects a genuine ambiguity in how broadly Google's brand entity recognition operates when the goal is blocking all known brand names.

The prior workaround and its limits

Before today's update, the primary mechanism for preventing AI Max from running on branded queries was a manually maintained brand exclusion list. That approach required advertisers to anticipate which brand names the campaign might otherwise match and add them proactively.

Analysis published in December 2025 found that existing brand inclusion and exclusion settings at the campaign level were missing many misspellings and word variations that still triggered matches. Negative keywords targeted specific text strings, but AI Max's intent inference could route around them by matching on inferred signals rather than literal query text.

The new unbranded-only toggle sidesteps that problem by operating at a higher level of abstraction. Rather than requiring advertisers to enumerate every brand variant they want to exclude, it delegates the brand recognition task to Google's entity index - a database maintained across its broader knowledge systems. Whether that index is comprehensive enough to cover all the edge cases practitioners encounter remains to be tested at scale.

AI Max controls: a growing stack

The branded search control arrives as the latest in a series of constraint mechanisms Google has layered onto AI Max since the format launched. The history of those additions tracks the sequence of advertiser objections.

Text guidelines, introduced in September 2025 and expanded globally in February 2026, gave advertisers control over the language AI Max generates in ad copy - up to 25 term exclusions and 40 messaging restrictions per campaign. Those controls govern what ads say, not which queries trigger them.

The Google Ads API v21, released on August 6, 2025, added programmatic support for AI Max, including the ai_max_setting.enable_ai_max field. Google Ads Editor 2.10, released on July 8, 2025, brought AI Max into the desktop application alongside campaign-level negative keyword support for Performance Max.

In April 2026, AI Brief launched, a feature that allows advertisers to supply natural-language instructions to guide how AI Max generates ads and selects searches - going beyond exclusions to allow positive framing of intent.

Taken together, the controls now available inside AI Max fall into four rough categories: query-level controls (negative keywords, branded search settings), creative-level controls (text guidelines, AI Brief), URL-level controls (final URL expansion exclusions, page feed rules), and reporting enhancements (AI Max expanded matches and expanded landing pages metrics).

The branded search toggle fits cleanly into the first category. It does not change how ad copy is generated, nor how landing pages are selected. Its scope is narrower: it determines which universe of searches the campaign is eligible to enter.

The broader context: AI Max as an expanding platform

Google declared AI Max out of beta on April 15, 2026, the same day it announced that Dynamic Search Ads would be retired and automatically upgraded to AI Max for Search campaigns in September 2026. That retirement is among the most structurally significant changes to Google's search advertising infrastructure in years and means advertisers who have not yet engaged with AI Max will be migrated to it regardless.

AI Max for Shopping campaigns launched on April 30, 2026, extending the same automation layer to standard Shopping ads with text customization drawn from Merchant Center feed attributes, final URL expansion, and format selection. Google has also merged travel ad formats into standard Search campaigns with AI Max, consolidating hotel, flight, and travel formats that previously operated through separate campaign types.

At Google Marketing Live 2026 on May 20, Vidhya Srinivasan, VP and GM of Google Ads and Commerce, confirmed that appearing in new AI Search ad formats - including Conversational Discovery ads and Highlighted Answers - requires AI Max for Search or Performance Max with text customization. The practical implication: advertisers who want access to inventory in AI Mode and AI Overviews will need to operate AI Max, not opt around it.

That trajectory makes control features like the branded search toggle more significant than they might appear in isolation. As AI Max becomes the mandatory infrastructure for accessing Google's newest ad surfaces, the granularity of what advertisers can constrain inside it determines how much of their campaign strategy survives the transition.

Attribution and budget implications

The dual-campaign structure - branded and non-branded running in parallel - is standard practice in paid search management precisely because the two serve different strategic functions. Branded campaigns defend against competitors bidding on an advertiser's own name, maintain impression share for navigational queries, and capture conversion data that is attributed cleanly to brand equity. Non-branded or prospecting campaigns operate in a different measurement frame, and blending the two data streams creates noise in any attribution model.

Independent testing documented by PPC Land in November 2025, covering more than 250 retail campaigns, found that AI Max delivered conversions at approximately 35% lower ROAS than traditional match types. A separate study cited at the time of the DSA retirement announcement revised Google's own headline figure downward from the original "14% more conversions" claim to an average of 7% more conversions when using the full AI Max feature suite - search term matching, text customization, and final URL expansion together - compared to search term matching alone.

Those performance figures are averages across a heterogeneous set of campaigns. An AI Max campaign that cannibalizes a significant share of branded traffic from a parallel branded campaign will produce misleading performance data in both directions: the prospecting campaign appears stronger than it is because branded queries convert at higher rates, and the branded campaign appears weaker because it is losing impressions. The unbranded-only toggle eliminates that confound at the source.

The practical advice circulating in the LinkedIn thread - that advertisers running separate branded campaigns should activate the unbranded-only setting - reflects this attribution logic as much as it reflects budget management. Branded clicks routed through a prospecting campaign that operates on broader targeting parameters and higher CPCs represent a real cost difference, but the downstream measurement distortion may be the more durable concern.

Rollout status

The feature appears to be in a limited rollout phase. Thomas Eccel posted about it on LinkedIn on May 29, 2026, and the screenshot he shared shows the setting within the AI Max campaign interface. Google has not published a formal announcement, and reactions in the LinkedIn thread suggest the control is not yet visible in all accounts. Several practitioners responding confirmed they had not yet seen the feature in their own setups.

This pattern - a feature first surfacing through practitioner screenshots before receiving formal documentation - has been consistent across AI Max's development history. Google Ads Editor 2.11 and successive API releases have typically followed initial interface rollouts by weeks or months.

Timeline

Summary

Who: Google, with the feature first publicly documented by Thomas Eccel, a paid search specialist, who shared a screenshot on LinkedIn on May 29, 2026.

What: Google is rolling out a native branded search control inside AI Max for Search campaigns. It offers three settings - show ads on all relevant searches, control branded searches with brand inclusions and exclusions, and show ads only on unbranded searches - giving advertisers a direct mechanism to prevent AI Max from matching on brand names, including competitor brands and the advertiser's own brand.

When: The feature became visible in some advertiser accounts around May 29, 2026, and is today being rolled out more broadly, though it has not yet received a formal announcement from Google.

Where: The control appears inside the AI Max campaign settings interface within Google Ads. It applies at the campaign level and draws on Google's own index of brand entities rather than relying solely on manually maintained exclusion lists.

Why: AI Max expands keyword targeting through intent inference rather than literal text matching, which has caused it to enter branded query auctions - including those for competitor brands and the advertiser's own brand - in ways that disrupt parallel branded campaigns, inflate CPCs, and distort attribution data. The new toggle provides a structural solution to that problem by confining AI Max to genuinely generic, non-branded traffic.