How Google's DSA to AI Max Upgrade Will Change PPC Workflows

Google to replace DSA with AI Max by 2026

Google announced on April 14, 2026 that Dynamic Search Ads are being retired, with AI Max for Search taking their place as the default system for automated search expansion. [6] Starting in September 2026, Google will automatically migrate all eligible DSA campaigns to AI Max, and new DSA campaign creation will be disabled across Google Ads, the Editor, and the API. [1] For PPC teams still running DSA as a keyword gap-filler, this is not a gradual sunset, it is a hard cutover with a firm deadline of September 30, 2026.

The migration also pulls in two other legacy automation features. Automatically Created Assets and campaign-level broad match settings are both being consolidated into AI Max, which means the upgrade touches more campaigns than just those explicitly built on DSA. [1] Google is not offering a parallel path where advertisers can keep the old tools running alongside the new system. After September, AI Max is the only option for this class of automated search behavior.

Google exited AI Max from beta at the same time as this announcement, citing adoption by hundreds of thousands of advertisers during the testing period. [6] That scale gives Google some cover for the forced migration, but it does not change the reality for advertisers who have built workflows around DSA’s crawl-and-match logic. Those workflows will need to be rebuilt from scratch, not just adjusted.

Why AI Max moves beyond website content

DSA’s core mechanism has always been straightforward: Google crawls your website, indexes the content, and uses that index to match search queries to relevant landing pages. It works well when your site is well-structured and your content is comprehensive, but it breaks down quickly when search behavior drifts away from what your pages actually say. Google’s argument for retiring DSA is that modern search patterns have become too unpredictable for a website-crawl model to keep up with. [1]

AI Max is designed to solve this by pulling in real-time intent signals from across Google’s search data, not just from your landing pages. [6] Where DSA asks “what does this website say?”, AI Max asks “what is this user actually trying to do, and which advertiser can best serve that intent?” That is a meaningful shift in how matching logic works, and it has real implications for which queries your ads appear on.

The practical consequence is that AI Max can match to long-tail queries that DSA would never have surfaced, because those queries have no direct equivalent on your website. [4] For advertisers with large product catalogs or service offerings that are hard to fully document in page copy, that expanded reach is genuinely useful. For advertisers who relied on DSA’s website-scoped matching as a form of implicit control, keeping ads tied to pages that actually exist, the shift to intent-based matching requires a different kind of guardrail strategy.

Google frames this as moving advertisers from manual, website-maintenance-dependent targeting to a scalable AI-driven growth model. [6] That framing is accurate, but it glosses over the fact that “scalable” and “controllable” are not the same thing. AI Max gives Google more latitude to match ads to queries that fall outside what advertisers have explicitly approved, which is a tradeoff worth naming clearly rather than burying in the announcement language.

How AI Max expands targeting and creative generation

AI Max operates through two distinct features that can be enabled or disabled independently. Search Term Matching expands reach beyond your keyword list using AI-driven intent signals. Asset Optimization dynamically tailors ad copy based on the context of each search. [7] Running both together is what Google counts as the “full AI Max feature suite,” and that combination is what drives the headline performance figure: an average of 7% more conversions or conversion value at a similar CPA or ROAS compared to using Search Term Matching alone. [6]

It is worth being precise about what that 7% figure actually measures. The comparison is full AI Max versus Search Term Matching as a standalone feature, not versus DSA or versus a manually managed keyword campaign. [4] Google is measuring the incremental value of adding Asset Optimization on top of Search Term Matching, which is a narrower claim than the headline implies. Advertisers comparing AI Max to their existing DSA performance should not assume a 7% lift is guaranteed, the baseline and methodology are different.

On the creative side, AI Max uses Gemini-powered models to generate ad copy that is more context-aware than what legacy ACA produced. [11] The system draws on advertiser-provided assets, website content, and real-time search context to assemble headlines and descriptions that fit the specific query, rather than pulling from a fixed asset pool. That is a genuine improvement over DSA’s headline generation, which was largely limited to page titles and category labels scraped from your site.

One agency that tested AI Max during beta reported turning off real-time ad customization because AI-generated headlines did not consistently match the specific intent of user searches, and they preferred keeping manual control over copy. [10] That experience reflects a real tension in the system: the AI optimizes for conversion probability at scale, but individual ad-query pairings can still produce copy that feels off-brand or misaligned. The toggles exist precisely because Google knows not every advertiser will trust the AI’s creative judgment unconditionally.

What new campaign controls AI Max provides advertisers

One of the more substantive differences between AI Max and DSA is the control layer that comes with it. DSA offered URL targeting and page feed options, but the levers for steering creative and matching behavior were limited. AI Max introduces a set of controls that let advertisers define the boundaries within which the AI operates, rather than simply turning automation on or off. [6]

Brand controls let advertisers specify how the system handles brand-related queries and creative, which matters for accounts where brand and non-brand traffic need to be managed separately. Location parameters allow geographic targeting to be defined at a level of precision that DSA did not support natively. Text guidelines let advertisers set messaging restrictions and exclusions, with limits of 25 term exclusions and 40 messaging restrictions per campaign. [12] These are not unlimited guardrails, but they are more structured than anything DSA offered.

URL controls have also been restructured. URL inclusions now operate at the ad-group level, giving more granular control over which pages are eligible for a given set of ads. URL exclusions operate at the campaign level, letting advertisers block entire sections of a site from being used as landing page candidates. [7] The Final URL Expansion toggle is the most consequential of these controls: when enabled, Google can dynamically select landing pages beyond the one specified in the ad, routing users to whatever page the system predicts will convert best. When disabled, the ad goes to the URL the advertiser specified. [1]

For accounts where landing page experience is tightly controlled, think regulated industries, legal or financial services, or any brand with strict compliance requirements, the Final URL Expansion toggle is the first thing to evaluate. Leaving it on by default means Google can send traffic to pages you have not explicitly reviewed for that campaign’s messaging, which creates compliance exposure that DSA’s more constrained matching logic did not generate at the same scale.

Search Term Matching and Asset Optimization can each be toggled independently at the ad-group or campaign level. [7] That modularity is useful for accounts that want the reach expansion of AI-driven query matching without handing over creative control, or vice versa. In practice, running Search Term Matching without Asset Optimization is probably the safer starting configuration for accounts migrating from DSA, since it preserves the familiar pattern of advertiser-controlled copy while expanding query coverage.

How the upgrade changes PPC campaign management

From what I have seen in how Google has progressively automated search over the past several years, the DSA-to-AI Max migration follows a consistent pattern: Google removes a tool that required advertisers to do explicit configuration work, replaces it with a system that handles that configuration automatically, and then offers a set of guardrails as a concession to advertiser control. The net effect is always the same, less direct management, more supervision of AI behavior.

For PPC managers, the shift changes where effort goes. DSA campaigns required ongoing maintenance of page feeds, URL targets, and negative keyword lists to keep matching behavior in bounds. AI Max requires a different kind of ongoing work: curating the asset library that feeds creative generation, maintaining exclusion lists that define what the AI cannot do, and auditing search term reports to catch matching behavior that drifts outside acceptable boundaries. [9] The workload does not disappear, it shifts from configuration to supervision.

Consolidating DSA, ACA, and campaign-level broad match into a single system also changes how accounts are structured. Advertisers who previously kept these features in separate campaigns to isolate their performance signals will need to rethink that architecture. AI Max is a unified system, and the controls it offers are designed to be set at the campaign or ad-group level within that system, not by splitting functionality across separate campaigns. [14]

The deeper implication is that strong inputs now drive performance more than tactical adjustments do. The quality of your creative assets, the precision of your exclusion lists, and the clarity of your text guidelines determine how well AI Max performs on your behalf. An account with a thin asset library and no exclusions will get broad, poorly targeted output. An account with well-structured assets, tight URL controls, and clear messaging restrictions will get output that is much closer to what a skilled human manager would have produced manually. The AI does not compensate for weak inputs, it amplifies whatever you give it. [15]

Accounts that have not yet run AI Max experiments should do so before September, not after. Google’s migration will be automatic for eligible campaigns, but going into it without any performance baseline means you will have no reference point for evaluating whether the migration helped or hurt. Running an AI Max experiment now, against your existing DSA campaigns, gives you data to inform how you configure the controls before the forced cutover arrives. [7] That window is closing faster than most migration timelines in Google Ads history.

Warsaw Streamer Łatwogang Breaks Guinness World Record for Charity Fundraising

How to Get Started in Digital Marketing

What Is llms.txt and Why It Matters for Your Content

What to Expect at Google Marketing Live 2026

Warsaw Streamer Łatwogang Breaks Guinness World Record for Charity Fundraising

What to Expect at Google Marketing Live 2026

Creating a Google Ads SKILL.MD for Claude

The Top E-mail Marketing Tools for Web Developers

What the 10.4% Growth in Global Ecommerce Means for Retailers

The Essential Digital Marketing Tool Stack for 2026

Adobe Summit 2026 Signals a Shift from AI Hype to Customer Action

Why AI Search Traffic Has a 5x Higher Conversion Rate

Dissecting the $306 Billion Global PPC Spend in 2026