Google AI Max: How It Works In Search Campaigns

Over the past several years, Google has gradually changed how Search campaigns work. Search advertising once depended almost entirely on manually selected keywords, advertiser-written ad copy, and landing pages chosen in advance. Advertisers picked exact terms, matched them to ad groups, and controlled almost everything.

As search behavior became more complex, Google introduced new automation features that helped advertisers reach additional queries, generate more ad variations, and direct users to more relevant pages on their websites.

Those changes happened gradually over several product releases. Dynamic Search Ads matched searches using website content instead of only keywords. Responsive Search Ads automatically combine headlines and descriptions. Broad Match evolved to understand search intent rather than relying only on exact words. Google later introduced Automatically Created Assets and Final URL expansion to generate additional ad copy and choose landing pages based on a user's search.

AI Max is the latest step in that progression.

The company announced AI Max in March 2025 as a system that brings several of those AI-powered capabilities together within existing Search campaigns instead of introducing another campaign type. The feature reached general availability and moved out of beta in May.

With AI Max now generally available, advertisers are asking practical questions before enabling it. They want to know what AI Max actually is, how it differs from Performance Max, whether it replaces Search campaigns, how it affects keyword targeting and landing pages, and how much control they still have over their campaigns.

This article answers the biggest questions advertisers have been asking about AI Max, explaining how the feature works, what has changed, and what it means for existing Search campaigns.

What Is Google AI Max?

Google's documentation describes AI Max as a comprehensive suite of targeting and creative features built around two main components, search term matching and asset optimization, that use AI to optimize ads in real time.

AI Max is not a new campaign type. It is an optional suite of AI-powered features that advertisers switch on within new or existing Search campaigns. Google is explicit on this point: AI Max is an optimization layer, not a standalone campaign format, and it sits on top of the Search campaign structure advertisers already use.

With AI Max, advertisers are not creating a new campaign type or migrating to Performance Max. Their campaign structure, existing keywords, and ad groups all stay intact.

What changes is what Google's systems are allowed to do with them.

AI Max uses Google's AI across three parts of a Search campaign: which search queries trigger an ad, what the ad copy says, and which landing page a click lands on. It layers these capabilities on top of the keywords, ad groups, and bid strategy an advertiser has already built. It does not replace the Search campaign underneath.

AI Max is a toggle that tells Google it can go beyond your exact keyword list to decide what triggers your ad, beyond your written headlines to decide what text appears, and beyond your chosen URL to decide where the click lands.

AI Max vs Performance Max

Since Google introduced AI Max, advertisers have viewed the system as either a replacement for Performance Max or an entirely new campaign type. That has led to one of the biggest questions since its launch: How is AI Max different from Performance Max?

Ginny Marvin, Google Ads' Product Liaison explained in a chat that AI Max for Search was built to bring feature parity with how Performance Max handles Search inventory. The key difference is where the two products run.

Although both products use AI, they serve different purposes.

Performance Max is a cross-channel campaign type that can serve a single campaign's budget across Search, Shopping, Display, YouTube, Gmail, Discover, and Maps simultaneously, with Google's AI making the channel allocation decisions on the advertiser's behalf.

By contrast, AI Max for Search campaigns serves exclusively on Google's Search Network, meaning the Search results page and Search Partner sites, if that setting is enabled. It keeps advertisers within the Search campaign structure they already understand and adds AI on top.

Control follows the same split. Standard Search campaigns give advertisers the most granular control: manual keywords, match types, and full search term visibility. AI Max sits in the middle. It keeps the keyword and ad group structure of Search campaigns but layers AI-driven expansion and creative generation on top, while preserving negative keywords, URL exclusions, and reporting that shows exactly which queries and assets drove results.

Performance Max gives up the most control in exchange for the broadest reach because it operates largely as a black box across non-Search inventory, with far less query-level transparency.

Advertisers can run both at once. Google's documentation addresses what happens when a single search query is eligible to match both an AI Max for Search campaign and a Performance Max campaign. Whichever campaign has the higher Ad Rank for that auction is the one that serves.

This means the two products compete against each other within the same account for overlapping queries. That is why agencies increasingly recommend monitoring cross-campaign overlap once both are active, particularly around branded terms.

How Does Google AI Max Actually Work?

AI Max operates through three linked mechanisms. All are switched on by default once the feature is enabled, and each can be controlled separately.

Search term matching

This feature expands an ad's reach beyond the exact keywords an advertiser has entered. It combines two technologies: broad match expansion and what Google calls "keywordless" matching, which serves ads based on contextual, semantic, and behavioral signals rather than a literal keyword the advertiser typed.

Traditionally, advertisers selected keywords they believed people would search for. While match types expanded over time, campaigns still depended heavily on keyword lists.

With AI Max enabled, Google's system can identify additional searches that express the same intent, even when those exact phrases do not appear in an advertiser's account. For example, an office furniture retailer may primarily target keywords related to ergonomic office chairs. Google says AI Max can also identify other relevant searches based on information from keywords, landing pages, creative assets, and other campaign signals when it determines that the advertiser's products are relevant.

Google's Help Center describes this as using a broader, more comprehensive understanding of the URLs across an advertiser's domain, built on a superset of the asset generation technology originally used in Dynamic Search Ads and Performance Max's Final URL expansion.

Search term matching is turned on by default at the campaign level but can be toggled off for individual ad groups.

Here's a Google video explaining how the feature work.

Text customization (asset generation)

This is Google's term for AI-generated ad copy. It generates headlines and descriptions by drawing on an advertiser's existing ads, landing pages, website content, and the keywords already in an ad group, then tailors that copy to match the specific intent behind an individual search.

This is the modern name for what Google previously called Automatically Created Assets.

Text customization is enabled by default and can be switched off. Disabling it also disables the next mechanism.

Final URL expansion

When active, Google's AI chooses which page on an advertiser's site to send a click to, rather than always using the landing page the advertiser designated.

For example, a search for "women's trail running shoes size 8" might land on a specific product page rather than a general running shoes category page, even if the ad was built around the category page.

In other words, a user searching for a specific product category may be sent directly to that category page instead of the website's homepage if Google believes it is the more relevant destination.

Google's documentation notes an important interaction here. Turning off text customization automatically disables Final URL expansion because the two features are linked. Advertisers who want to keep a static, advertiser-chosen landing page throughout a campaign need to disable Final URL expansion specifically, not just review the setting.

Does Google rewrite existing ads outright? Not in the sense of altering an advertiser's manually written Responsive Search Ad assets. Text customization generates new asset combinations for specific queries, but advertisers can pin existing headlines and descriptions to prevent Google from replacing them. Any manually written copy also remains unaffected by the text guideline controls covered below.

Where Final URL expansion is active, however, Google may bypass pinned assets when it determines that a different URL and headline combination better matches a user's intent, unless Final URL expansion itself is turned off.

Should Advertisers Enable Google AI Max?

Google's Help Center addresses this question directly. Its AI Max overview page states that its purpose is to explain how AI Max works "and when you should use it," and ties the case for enabling it to its published benchmark: advertisers activating AI Max typically see 14% more conversions or conversion value at a similar CPA or ROAS, based on Google's internal 2025 data for non-retail advertisers.

That 14% figure explicitly does not apply to retail advertisers. Google's own documentation excludes ecommerce and retail accounts from the benchmark, which is worth considering before assuming the same case for enabling AI Max carries over to a retail account.

There's also a technical prerequisite. Google's setup documentation states that search term matching, Final URL expansion, and text customization all require a conversion-based Smart Bidding strategy, meaning Maximize Conversions, Maximize Conversion Value, Target CPA, or Target ROAS, to function effectively. 

Accounts still running Manual CPC or a bid strategy that is not optimizing for conversions are unlikely to get the full value from AI Max.

On where advertisers should be most cautious, Search Engine Land's reporting points specifically to brand campaigns. One contributor's account audits found that budget constraints, weak landing page alignment, poor account structure, and outdated query management were usually the real bottlenecks limiting generic, non-brand campaigns. That suggests the strongest case for AI Max is often in generic campaigns rather than in brand campaigns that are already performing well.

Google's own briefings to advertisers also acknowledge brand safety concerns directly, including the risk of AI-generated copy misattributing claims from one product to another. That is one reason some advertisers running claims-heavy or compliance-sensitive messaging have been slower to adopt AI Max broadly.

Google's product team has also clarified that keyword prioritization is unaffected by enabling AI Max. If a user's query matches an eligible Search keyword of any match type, that keyword-based match takes priority over AI Max's keywordless or broad match expansions. Ad group theming still matters strategically once AI Max is enabled because search term matching draws on the keywords, assets, and landing pages already present in an ad group to discover new queries.

The practical way to think about Google's published performance figures is to treat them as an upper-bound benchmark based on favorable internal data, not as a guarantee. Run AI Max as a monitored test against a documented baseline rather than assuming the published percentages will apply to every account.

What Control Do Advertisers Still Have?

Google's guidance says AI Max does not take over campaign management. While AI Max automates more decisions during ad serving, advertisers continue to define the strategic boundaries within which those systems operate, including budgets, targeting, exclusions, brand controls, and campaign objectives.

Keywords still work: Exact match and phrase match keywords remain fully functional within an AI Max campaign. AI Max search term matching expands beyond them; it does not replace them.

Negative keywords and search term exclusions: Standard negative keyword lists still apply and remain the primary tool for limiting AI Max's query expansion. Agencies consistently recommend auditing and strengthening negative keyword lists before enabling AI Max, not afterward.

Brand and URL exclusions: Advertisers can set brand inclusions and exclusions at both the campaign and ad group levels, and they can exclude specific URLs from ever being used by Final URL expansion.

Last month, Google began testing a more direct "unbranded only" toggle for AI Max campaigns. It is intended to give advertisers a simpler way to prevent the feature from serving ads on their own branded queries without manually listing every brand variation and misspelling as negative keywords.

Text guidelines and term exclusions: Text guidelines let advertisers create natural-language rules that constrain what AI-generated ad copy can say, including tone, compliance language, and required disclaimers.

Alongside those guidelines, advertisers can set up to 25 hard term exclusions per campaign, meaning specific words or phrases that the AI is never allowed to use in generated copy, as well as up to 40 broader messaging restrictions.

Google says these controls govern only what generated copy says, not which search queries trigger an ad. Query matching is controlled separately through keywords, match types, negative keywords, and search term matching settings.

Locations of interest: This AI Max-specific control lets advertisers target users based on geographic intent at the ad group level, even for keywordless matches.

Budgets: Budget settings and bid strategies remain standard Search campaign controls. AI Max does not change how advertisers manage spending.

Reporting: The Search terms report now includes a distinct "AI Max" match type with a Source column explaining why a match occurred. The Keywords report shows AI Max's contribution, and the Landing pages report includes a "Selected by" column so advertisers can see when Final URL expansion selected a page.

Asset removal: Advertisers can remove individual AI-generated headlines or descriptions that do not fit brand standards through the asset report. However, bulk removal of Google-customized text assets currently requires either the Google Ads API or Google Ads Editor rather than the standard web interface.

AI Brief: AI Brief lets advertisers give AI Max natural-language instructions that go beyond exclusions, allowing positive guidance on tone and messaging instead of relying solely on restrictions.

What advertisers cannot fully control is the bidding logic, since AI Max works within an existing bid strategy such as Target CPA or Maximize Conversions, rather than changing it. Nor can they fully control the underlying decision about which specific query, from a near-infinite long tail, gets matched to an ad on a given day. These controls serve as guardrails, not a return to fully manual targeting.

Does AI Max Replace Keywords?

No. AI Max complements and expands keyword targeting rather than replacing it. Keyword research remains relevant because AI Max's search term matching still uses an advertiser's existing keywords and ad group structure as its foundation. What changes is how much incremental query volume beyond that foundation gets captured automatically through broad match expansion and keywordless matching, rather than requiring advertisers to anticipate and add every relevant keyword manually.

Whether broad match becomes the practical default is a fair concern. Because search term matching is turned on by default when AI Max is enabled, and because it relies heavily on broad match-style expansion, accounts that do not actively manage negative keywords can end up functioning much more like broad match than advertisers intended, even if their keyword lists are still built around exact and phrase match.

This is one of the most common criticisms raised by paid search analysts. AI Max can quietly stretch match-type behavior well beyond what the keyword list itself implies, making it harder to understand exactly what is driving results without paying close attention to the reporting discussed below.

How Can Advertisers Enable AI Max?

AI Max is enabled from within a Search campaign's settings, in the Asset optimization panel, where advertisers will find toggles for search term matching, text customization, and Final URL expansion. Text customization and Final URL expansion are both enabled by default once AI Max is switched on for a campaign, although each can be turned off individually. However, disabling text customization also disables Final URL expansion because the two features are linked.

For new campaigns, advertisers can build directly with AI Max enabled from the start. AI Max appears as a dedicated step during campaign creation. Advertisers can toggle it on and then choose which features to activate: search term matching, text customization, and Final URL expansion. Ad group-level settings are applied afterward within the Keywords and ads section. Google’s AI Max set up page explains further how to enable AI Max in Google Ads.

On the question of whether existing campaigns can be upgraded, the answer is yes. Increasingly, however, this is becoming less of a choice. 

Google in April said that campaigns using Automatically Created Assets and campaign-level broad match settings would begin auto-upgrading to AI Max starting in September 2026. The broader Dynamic Search Ads sunset and auto-upgrade will begin in February 2027 after Google extended the original timeline to protect advertiser performance during the transition.

Google says it will attempt to mirror an account's existing settings during the automatic conversion to preserve performance stability. However, advertisers who want greater control over the migration timeline and configuration are advised to upgrade voluntarily before the automatic conversion window opens.

Conclusion

AI Max for Search campaigns is no longer a feature advertisers can treat as optional testing. Google has confirmed that Dynamic Search Ads will be automatically converted to AI Max on a fixed timeline. For advertisers running those campaign types, the only real decision left is whether to migrate on their own terms now, with time to configure controls and establish a performance baseline, or wait for Google's automatic conversion.

The core tradeoff has not changed since AI Max launched. AI Max expands a Search campaign's reach into queries and creative variations that an advertiser's manual keyword list would otherwise miss, and Google's own data estimates a measurable upside for many non-retail advertisers. At the same time, it hands more of the query-matching and copy-generation process to Google's AI than a standard Search campaign ever did. 

The company has also introduced a distinct AI Max match type in the Search terms report so advertisers could isolate its performance from broad match. It later rolled out a unified report that brings together search terms, ad headlines, and landing pages in a single view.

For most advertisers, the recommended approach is the one reflected in Google's own documentation and setup guidance: confirm that conversion tracking and bidding strategy are ready, enable AI Max on generic, lower-risk campaigns first, and use the reporting Google has built, specifically the AI Max match type, the Source column, and the Landing pages report, to verify that the feature is performing as expected before expanding it further.

AI Max is becoming the default state of Google Search advertising. How much control advertisers retain within that default depends on how deliberately they configure it.