A campaign type you have run for years is being retired.
Google AI Max for Search is Google’s AI-powered replacement for Dynamic Search Ads: a campaign setting that uses AI to find new queries, rewrite ad text, and match searchers to the right landing page, all within your Search campaigns. It left beta in April 2026, and Google has confirmed that Dynamic Search Ads and other legacy features will be automatically upgraded to it. For anyone running Search campaigns, this is not a feature to evaluate at leisure. It is a migration with a deadline, and the advertisers who plan for it will keep far more control than the ones who wait for Google to flip the switch.
The good news is that the deadline moved. In June 2026, Google pushed the automatic migration of Dynamic Search Ads from September 2026 to February 2027 and restored the ability to create new Dynamic Search Ads campaigns. That gives you a real runway to test, compare, and migrate on your own terms, if you use it. This guide walks through what AI Max actually is, what the timeline requires, what the performance data says once you strip away the marketing, and how to move without handing your account over blind.
What is Google AI Max for Search?
Google AI Max for Search is a setting inside a standard Search campaign that turns on three AI features at once: search term matching, which finds relevant queries beyond your keyword list; text customization, which generates and adapts headlines and descriptions to each search; and final URL expansion, which sends people to the most relevant page on your site rather than the one URL you picked. Turn it on and Google’s models take over the parts of a Search campaign that Dynamic Search Ads used to handle, but with broader, real-time intent signals instead of just your landing-page content.
It is easiest to understand as the next generation of Dynamic Search Ads. For years, advertisers used Dynamic Search Ads as a catch-all to capture traffic outside their keyword lists, generating headlines and landing-page matches automatically from website content. AI Max does the same job, then adds two things Dynamic Search Ads never had: richer signals for finding untapped queries, and a set of steering controls (brand controls, location controls, and text guidelines) that let you tell the AI what it may and may not do. That control layer is the whole story for advertisers who have been burned by automation before, and we will come back to it.
Is AI Max really replacing Dynamic Search Ads?
Yes. Google has confirmed that Dynamic Search Ads is being retired and folded into AI Max for Search, along with two other legacy features: automatically created assets and the campaign-level broad match setting. This is not an optional test. It is a platform-wide migration, and Dynamic Search Ads (one of Google’s longest-running Search formats) is on the way out.
The reason Google gives is that searches have gotten more complex and less predictable, and a system built only on your landing-page content can no longer keep up with how people phrase things. Whether or not you buy that framing, the practical takeaway is the same: the campaign type is going away, and the replacement is on by default once the upgrade runs. Demand for information about the change is real and rising: Semrush keyword data (US, July 2026) shows “dynamic search ads” drawing about 720 searches a month at a $16.50 cost-per-click, and “google ai max” also pulling roughly 720 a month, a sign that advertisers are actively trying to get ahead of this.
When does the DSA-to-AI Max upgrade actually happen?
The automatic upgrade of Dynamic Search Ads campaigns now begins in February 2027, not September 2026 as Google first announced. In June 2026, after advertiser feedback, Google delayed the forced migration by five months and reopened Dynamic Search Ads campaign creation, which it had planned to close. That gives you the second half of 2026 and the start of 2027 to prepare instead of scrambling.
Here is the sequence that matters, according to Google’s own announcements and Search Engine Land’s June 2026 reporting:
- June 15, 2026: the ability to create new Dynamic Search Ads campaigns returns, and voluntary migration tools are available to port your settings and history into AI Max.
- June 2026 – January 2027: the extended testing window — run AI Max side by side with your existing setup and compare on your own numbers.
- January 2027: new Dynamic Search Ads creation ends for good.
- February 2027: automatic migration begins for any remaining Dynamic Search Ads campaigns.
One caveat worth flagging: the February 2027 extension applies specifically to Dynamic Search Ads. Google’s original plan also swept in automatically created assets and the campaign-level broad match setting, and those two were part of the earlier September 2026 upgrade wave. If your account relies on either of those, do not assume the Dynamic Search Ads delay covers you: check each campaign’s settings rather than the headline date.
Does AI Max actually improve performance?
Sometimes, and by less than the headline suggests once you leave Google’s preferred test conditions. Google’s own figure is a 7% average lift in conversions or conversion value at a similar cost-per-acquisition when you run the full AI Max feature suite versus search term matching alone, and that number, per Google’s footnote, comes from internal data on non-retail advertisers. That is a specific, narrow comparison, not a promise of 7% more results than your current account produces.
Independent testing tells a more sobering story. Smarter Ecommerce, led by Mike Ryan, analyzed more than 250 retail Search campaigns running AI Max and found a median revenue lift of 13%, close to Google’s optimistic framing, paired with a median cost-per-acquisition increase of 16%. The spread was wide: return-on-ad-spend outcomes ranged from plus 42% to minus 35%, and only 22% of campaigns landed anywhere near their original efficiency targets. In other words, AI Max reliably buys you more volume, but it often buys that volume at a worse rate, and the outcome for any single account is far from guaranteed.
Ryan’s blunt summary was that flipping AI Max on is “essentially a coin toss.”
The lesson is not “avoid AI Max.” It is “do not treat Google’s benchmark as your forecast.” Extra volume at a higher cost-per-acquisition is a good trade for some businesses and a terrible one for others, and the only way to know which you are is to test against your own numbers before the forced upgrade decides for you. Measuring that honestly (volume and efficiency together, not one in isolation) is the same discipline we bring to measuring the true ROI of AI-era search.
How is AI Max different from Performance Max?
AI Max lives inside a Search campaign and works on search intent only; Performance Max spans every Google surface (Search, Shopping, YouTube, Display, Gmail, Maps) and decides where your budget goes with far less visibility for you. That distinction is the whole point. With AI Max you are still running a Search campaign, still see your search terms, and still keep keyword-level structure and controls.
For advertisers who want AI to expand their reach without giving up the reporting and steering that Search campaigns provide, AI Max is the more controllable option of the two. It is closer to “Dynamic Search Ads with better signals and real guardrails” than to “hand Google everything and hope.” If you have wrestled with the reporting gaps and budget opacity of Performance Max, you will recognize the trade-offs immediately. We covered them in depth in our look at escaping the Performance Max black box. The practical guidance: keep AI Max and Performance Max in separate campaigns with clear roles, and do not let the automatic upgrade quietly blur the line between a Search strategy and an everywhere-at-once strategy.
How should you migrate from DSA to AI Max?
Migrate on your schedule, with the controls turned on, before February 2027 makes the choice for you. The advertisers who lose the most in this transition are the ones who let the automatic upgrade run on default settings and only notice the cost-per-acquisition drift weeks later. A deliberate migration takes a few hours and protects your account.
- Audit your Dynamic Search Ads campaigns first. List every campaign still using Dynamic Search Ads, automatically created assets, or campaign-level broad match. Note which drive real conversions and which are coasting — you do not want to carry dead weight into a new system.
- Run AI Max as an experiment, not a swap. Use Google’s one-click experiment tools to run AI Max against your current setup on the same campaign, and let it gather enough data to judge on cost-per-acquisition and return-on-ad-spend, not just raw conversion count. Given the wide performance spread in independent testing, your own numbers are the only ones that matter.
- Turn on the steering controls from day one. Set brand controls so your ads do not show on irrelevant or competitor-brand queries, use location controls to hold geographic relevance, and write text guidelines so the AI’s generated copy stays on-message. These controls are the single biggest difference between AI Max and the automation that preceded it — skipping them is how you get the “coin toss” outcome.
- Lock down URL expansion. Keep final URL expansion pointed at the pages you want to rank as destinations, and exclude the ones that should never be an ad landing page — checkout steps, thin pages, policy pages.
- Migrate the winners manually before the deadline. For campaigns that pass the experiment, port them yourself using Google’s upgrade tools so your settings and history carry over cleanly, rather than waiting for the automatic upgrade to apply generic defaults.
Done this way, the deadline stops being a threat and becomes a prompt to tighten campaigns you were going to have to revisit anyway. If paid search is a meaningful part of your pipeline and you would rather not run this migration blind, that is exactly the kind of work our paid search management is built around.
The controls are the campaign: The single most repeated finding in independent AI Max testing is how much the outcome swings on the settings. Turned on raw, results ranged from a 42% gain to a 35% loss on return-on-ad-spend, and only about one campaign in five stayed near its target. Brand exclusions, location controls, text guidelines, and URL rules are not optional fine-tuning. They are the difference between AI Max as a growth lever and AI Max as a budget leak.
What should advertisers do right now?
Start testing before the end of 2026, so the February 2027 upgrade lands on campaigns you have already tuned rather than ones running on autopilot. The extension Google granted is not a reason to relax. It is the window to do this properly. Pull your list of Dynamic Search Ads, automatically created assets, and broad match campaigns this quarter, launch one AI Max experiment against a campaign you understand well, and judge it on efficiency and volume together.
What you are really buying with an early migration is control: control over which queries you expand into, which pages you send traffic to, and whether the extra reach comes at a cost-per-acquisition your business can absorb. Wait for the automatic upgrade and Google makes all of those decisions with generic defaults. Move first, and you keep them.
Google AI Max for Search is the confirmed replacement for Dynamic Search Ads, and remaining DSA campaigns will be auto-upgraded to it in February 2027: five months later than Google first planned. The performance case is real but overstated: Google cites a 7% average lift for non-retail advertisers, while independent testing of 250+ campaigns found a 13% median revenue gain paired with a 16% higher cost-per-acquisition and outcomes ranging from plus 42% to minus 35%. The deciding factor is the controls. Audit your legacy campaigns now, run AI Max as a side-by-side experiment judged on your own efficiency numbers, turn on brand, location, text, and URL controls from the start, and migrate your winners manually before the deadline decides for you.
Facing the AI Max migration and want it handled right?
We’ll audit your Dynamic Search Ads campaigns, run controlled AI Max experiments against your real numbers, and migrate the winners with the guardrails on, so the February 2027 deadline works for you, not against you. Free, no obligation.
Sources
- Google Ads (Brandon Ervin, Director of Product Management), “We’re upgrading Dynamic Search Ads to AI Max,” Keyword blog, April 15, 2026: AI Max for Search leaving beta; average 7% more conversions or conversion value at similar CPA/ROAS from the full feature suite (search term matching, text customization, final URL expansion) vs. search term matching alone (Google internal data, 2026, non-retail advertisers); confirmed automatic upgrade of Dynamic Search Ads, automatically created assets, and campaign-level broad match; steering controls for brand, location, and text
- Search Engine Land (Anu Adegbola), “Google delays Dynamic Search Ads migration to AI Max,” June 12, 2026: automatic DSA migration pushed from September 2026 to February 2027; DSA campaign creation restored June 15, 2026; new DSA creation ends January 2027; automatic migration begins February 2027
- Smarter Ecommerce (Mike Ryan), AI Max analysis of 250+ retail Search campaigns, reported by Search Engine Land, 2026: median revenue lift of 13% with a median CPA increase of 16%; return-on-ad-spend outcomes ranging from +42% to -35%; only about 22% of campaigns near their original efficiency targets
- Smarter Ecommerce, AI Max match-type expansion analysis (1M+ impressions), 2026: query expansion roughly 80% exact match, 20% phrase match, under 1% broad match; Google’s headline uplift figure excludes retail advertisers
- Search Engine Journal, “What SMEC’s Data Reveals About AI Max Performance,” 2026: corroboration of the Smarter Ecommerce findings on volume gains paired with efficiency costs
- Semrush keyword data (US), July 2026: “google ai max” ~720 searches/mo; “dynamic search ads” ~720/mo at $16.50 CPC; “ai max for search” ~260/mo; “ai max for search campaigns” ~170/mo at $16.69 CPC