Update — April 2026: AI Max for Search is now GA
Google pulled AI Max for Search out of beta on 15 April 2026 after eleven months of open testing. In the same announcement, Google said Dynamic Search Ads, Automatically Created Assets, and campaign-level broad match will auto-upgrade to AI Max in September 2026, and new DSA creation will be blocked across the Google Ads UI, Google Ads Editor and the Google Ads API from that date. Voluntary upgrade tools began rolling out the week of 15 April. This post has been rewritten around that news — the old framing referenced a “Performance Max 2.0” and a pilot “transparency hub” that Google never actually branded or shipped under those names.
Brandon Ervin, Director of Product Management at Google Ads, announced that AI Max for Search has reached general availability in the same post that confirmed the DSA migration timeline.
What actually shipped on 15 April 2026
Three things are now live or confirmed in the GA release:
- AI Max for Search as a standalone, toggleable campaign layer. It's not a new campaign type — it's a setting you enable on existing Search campaigns, bundling search term matching, text customisation and final URL expansion.
- The Search terms and landing pages from AI Max report. Available from the Search terms dropdown, it shows which keyword-less queries triggered ads, which AI-generated headlines ran, and which landing pages they routed to — the transparency layer that was missing in Performance Max.
- A dated sunset for DSA. September 2026 is the hard deadline for auto-migration. If you rely on DSA today, you have roughly five months to audit landing-page quality and negative keyword coverage before the upgrade is imposed.
Notice what is not in that list: the “Performance Max 2.0” product name from the original 2025 post. It was never Google's branding. The real shift has always been called AI Max for Search — a distinct product from Performance Max, which is itself being upgraded separately with Gemini reasoning models.
The DSA sunset: your September 2026 deadline
For the agencies and in-house teams still running Dynamic Search Ads at scale, the September 2026 migration is the single most important date on the 2026 paid-search calendar. AI Max inherits DSA's keyword-less targeting logic but layers on generative text customisation and final URL expansion, which means the same campaign that used to ride off a tightly curated set of landing pages will now have Google's models writing new headline variants on the fly and, in some cases, sending clicks to URLs the advertiser never manually specified. The controls that survive — negative keywords, brand exclusion lists, location-of-interest targeting — become the load-bearing guardrails of the whole account. If those aren't clean, the upgrade will expose it.
What the real 2026 numbers say about AI Max performance
This is where 2025 marketing blogs got overenthusiastic. The honest 2026 picture:
- Google's own data: AI Max campaigns see an average of 7% more conversions or conversion value at a similar CPA/ROAS when using the full feature suite (search term matching + text customisation + final URL expansion) compared to search-term matching alone. Advertisers activating AI Max typically see 14% more conversions or conversion value; accounts that were previously dominated by exact and phrase keywords report up to 27% uplift.
- Independent data: A Monks Agency multi-account analysis of roughly 30,000 AI Max search terms found that 99% of impressions had zero conversions. A LinkedIn poll of PPC professionals found only 16% reported good performance, while 84% reported neutral or negative results.
- The synthesis: 10–20% conversion improvement at similar CPA is a realistic target for well-prepared accounts — accounts with clean conversion tracking, automated bidding, exhaustive negative keyword lists and enough budget to give the model statistically meaningful signal. The gap between Google's best-case numbers and independent worst-case numbers is almost entirely explained by account readiness.
The original 2025 version of this article quoted a “17% ROAS pilot” stat that was not verifiably sourced and has been removed.
Performance Max in 2026: Gemini, Veo, Imagen 2 and the channel timeline
While AI Max for Search is the headline news, Performance Max got its own set of 2026 upgrades that are worth mapping alongside it because the two products now share most of their generative stack:
- Gemini models now power long headline generation and sitelink generation inside Performance Max, using the same reasoning capability as Gemini-powered Search features.
- Imagen 2 — adapted for Google Ads — is rolling out for lifestyle imagery generation, including images of people in scenes rather than product-on-background.
- Veo video generation is now live in Asset Studio for Performance Max video assets.
- AI voice-over on Performance Max videos began rolling out with an opt-out deadline of 20 March 2026 — if you missed that, your PMax video ads may already have Google-generated narration.
- Channel performance timeline launched in early April 2026, giving advertisers a visual graph of how Search, YouTube, Display, Discover, Gmail and Maps each contributed — the single biggest visibility upgrade PMax has ever shipped.
- Brand safety controls — Google Ads Editor 2.12 (March 2026) now supports up to 25 term exclusions and 40 messaging restrictions per campaign for both Search and Performance Max, the direct response to marketers who found Final URL Expansion and automated asset generation producing off-brand or legally risky copy.
Taken together, AI Max + the PMax 2026 updates mean a modern Google Ads account is running AI-generated headlines, AI-generated images, AI-generated voice-over and AI-selected landing pages across every placement Google sells. Creative review is no longer a one-time sign-off — it's a continuous governance job.
A tactical playbook for the AI Max transition
Five moves before September:
- Audit conversion tracking and offline conversion imports. AI Max steers on signal; if your signal is dirty, it will steer badly.
- Rebuild your negative keyword lists at the account level. Keyword-less targeting means Google will find queries you never thought to bid on. Negatives are now your main lever.
- Enable the AI Max search terms and landing pages report on day one. Don't wait. You need a baseline of what's actually triggering.
- Use the Google Ads Editor 2.12 brand guidelines. Specify term exclusions and messaging restrictions before you enable text customisation. Assume the model will try every permutation.
- Test AI Max on one campaign before September. A controlled A/B with matched budgets is the only way to know whether your account is in the 16% that sees lift or the 84% that doesn't.
What this means for marketers
The larger story is that the unit of execution in paid search has changed. For fifteen years, the unit was the keyword — the thing you bid on, the thing you wrote an ad for, the thing you optimised. AI Max replaces that unit with a bundle of inputs: landing pages, existing ad copy, assets, brand guidelines, conversion signals, negatives. Google's models decide, in real time, how to combine those inputs into a served ad for a query your spreadsheet has never seen. This makes the marketer's job less about tuning individual ads and more about governing the inputs — making sure the landing pages, the brand rules, the creative library and the conversion data all stay coherent and current across every channel, not just search. That is not a Google Ads problem. That is an operating-model problem, and no single point tool solves it. Anjin was built for exactly this governance layer.
Anjin: the Marketing Operating System for the post-keyword era
Anjin is the Marketing Operating System built for exactly this shift. It holds your brand guidelines, your approved messaging, your creative library, your audience definitions and your channel performance in a single coherent layer — so when AI Max for Search, Performance Max, Meta Advantage+, TikTok Smart+ or the next generative ad system goes looking for inputs, every one of them draws from the same governed source of truth. You don't spin up a new workflow for each platform. You don't rebuild the same brand safety rules in five different editors. The Marketing OS runs the governance job continuously, and the ad platforms consume it. For agencies, that turns the AI Max migration from a fire-drill into a configuration step. For in-house teams, it turns an unmanageable September deadline into a normal Tuesday.
The £888 Lifetime License — Offer Closing Soon
Lifetime access to Anjin for a one-time payment of £888. Not a subscription. Not a seat. Not a trial. One payment, unlimited use, for as long as Anjin exists.
The average marketing team spends £888 in about three working days on tooling, freelancers and coordination software. You're buying the platform that replaces most of it — once.
This price will not be offered again once we close our early-access cohort.
Claim your £888 Anjin lifetime license →Founders, agency owners and in-house marketers — this is how you run marketing at AI speed without the team, the burn, or another year of waiting.
Sources: Google Blog, Google Ads Help, Yahoo Finance, AdWeek, MediaPost, ALM Corp, PPC.land, Google Business
