On August 14, 2026, Google announced that advertisers still running legacy search campaign formats — Dynamic Search Ads (DSA) and campaign-level Broad Match paired with Automatically Created Assets (ACA) — are being force-migrated to its newer "AI Max for Search" format, with no opt-out. The forced migration window runs from September 1 to 30, 2026 — meaning it's happening right now: some accounts have already flipped, and the rest will be converted automatically before the month is out. Creating new legacy-type campaigns has been impossible since August 3. Advertisers technically had the option to upgrade voluntarily, via in-account banners and tools, but starting in September Google is simply doing it for them, carrying over historical settings, brand exclusions and inclusions into the new format.
What actually changes inside the campaign
AI Max isn't a cosmetic rename. The bundle ships three things at once: automated, broadened search-term matching (AI decides which queries trigger the ad, beyond the formal keyword list), AI-written headlines and ad copy, and — the part that matters most — final URL expansion, where the system can send a click to a different page on the site than the one the advertiser specified, if it judges another page more relevant. Google's own claim is that "AI Max for Search campaigns see an average of 7% more conversions," but that figure comes from the company's own comparison against campaigns without AI Max, not an independent audit — it's worth treating as marketing copy rather than a guarantee for any specific account.
Who should be logging into Google Ads right now
Picture a small window-and-door installation business in Porto, running an €800-a-month ad budget with no dedicated PPC person — the owner manages campaigns himself in the evenings. He had one DSA campaign covering the entire product catalog, plus a few Broad Match campaigns set up two years ago and never touched since. In September those campaigns automatically become AI Max: the system starts deciding on its own which queries to show up for, writing its own ad headlines, and choosing which page on the site to send traffic to. If he doesn't have an up-to-date negative-keyword list and doesn't review brand controls, part of his budget could go, within the first few weeks, to irrelevant search terms — or send visitors to a generic catalog page instead of the specific product page, with worse conversion — and he'll likely only notice a month later, when the bill arrives.
This isn't a hypothetical — it's exactly what PPC professionals have been warning about for years with every new "smart" Google Ads format: broadened automated matching and auto-generated assets only work well with regular monitoring, search-term report reviews, and constant negative-keyword upkeep. Without that, they quietly drive up spend by picking up adjacent or loosely related queries that technically "convert" but don't bring profitable customers. For a large advertiser with an analytics team, that's a manageable risk. For a small business with nobody reviewing reports weekly, it's a direct path to overspend with no immediate visible warning sign.
One practical detail worth checking now: if the account is managed via the Google Ads API — custom bidding scripts, CRM integrations, or BI dashboards — it's worth confirming the API version is current. Support for older versions and the fields tied to legacy features is set to end around September 2027, and it's better to test compatibility now than to untangle a broken integration after the fact.
Behind all of this sits a broader pattern: Google is systematically stripping out manual levers for advertisers and replacing them with models that decide, on their own, who sees what and where. That can genuinely improve results for accounts that never had the resources for fine-grained tuning in the first place. But it also means the difference between "a well-configured AI" and "an AI quietly burning budget unsupervised" now depends entirely on whether someone is watching the reports closely after the migration — not on how smart the system itself is.
At Dayava, this is exactly the kind of territory where process automation should work for a business, not against it: setting up ad-spend monitoring, anomaly alerts, and regular reviews of negative keywords and brand controls is automation too — just aimed at keeping someone else's automation in check.
Source: Search Engine Land, "Google sets AI Max migration timeline for Search campaigns"
If a similar solution for your business sounds interesting — leave a request at dayava.pt/contactos/.