Google has set a date for a significant change that many paid search teams anticipated but hadn't fully prepared for. Starting September 1, 2026, Google will automatically upgrade eligible Search campaigns to AI Max if they utilize campaign-level broad match and/or Automatically Created Assets. This rollout will occur gradually throughout September, as reported by Search Engine Land and other industry sources.
AI Max is not just another campaign type to ignore; it’s an automation layer within Search campaigns. Once implemented, management will shift from manual keyword control to focusing on inputs: assets, negatives, URL rules, brand constraints, and measurement discipline. This represents a different workflow and a new potential for failure.
Moreover, on August 3, 2026, Google ceased allowing new legacy ACA and campaign-level broad match configurations in the Google Ads UI, Ads Editor, and API, signaling that the off-ramp is closing.
The real issue isn’t migration. It’s what changes after it.
For B2B growth leaders, the key concern isn't merely Google's addition of AI. The operational challenge lies in how AI Max can expand query matching and ad generation, impacting lead quality, funnel efficiency, and creative control. If your CPA remains stable while qualified pipeline diminishes, the dashboard may appear fine until Sales raises concerns. By then, the account may have already drifted.
Google claims that advertisers activating AI Max typically see about 14% more conversions or conversion value at similar CPA or ROAS, with potential lifts of up to 27% for campaigns relying heavily on exact and phrase match. While these figures are noteworthy, they are still Google’s claims. Independent commentary suggests that outcomes can vary widely by account, indicating potential upside but cautioning against assuming guaranteed uplift.
This necessitates treating the transition as a controlled restructuring project rather than a simple settings cleanup. Successful teams will enter September with established baselines, exclusions, and a readout plan. Those unprepared will find themselves in Q4 debating whether declines in traffic, quality shifts, or reporting gaps stem from AI Max, seasonality, or sales follow-up.
Run the test before Google runs it for you
Here’s a quick action plan: identify all Search campaigns using campaign-level broad match or ACA, export your baseline, and set up a 50/50 Google Ads Experiment to compare AI Max on versus off within the existing campaign structure. Experts suggest using sufficient volume to yield results, with one source recommending about €50 per day over 4 to 6 weeks.
The hypothesis: enabling AI Max in a controlled experiment will either maintain or increase conversion volume while keeping qualified pipeline efficiency within acceptable limits, as broader matching and text customization capture incremental intent.
Success is defined as a stable or improved qualified conversion rate and downstream efficiency, with guardrails including search term relevance, landing-page engagement, and brand-safe asset output. A stop-loss condition would be a sustained decline in lead quality or a significant rise in irrelevant query volume before the test concludes.
What to measure: export conversion data, search terms, and landing-page metrics before the test begins. If API support or workflow coverage lags during the transition, these exports will serve as your control record. Remember, platform attribution is directional, not definitive. If the account is large enough, pair the experiment with a holdout logic at the campaign cluster level to compare movements in qualified stages, not just top-funnel form fills.
Control moves upstream: negatives, URLs, assets
Once AI Max is active, especially in campaigns tied to ACA, Google can enable search term matching and text customization by default. This can cause unease among B2B teams, as increased automation can widen reach but also the gap between a click and a sales-accepted opportunity.
The practical response is governance: tighten your negative keyword strategy, define URL inclusion and exclusion rules, and establish brand and creative guardrails in writing, particularly if your category has compliance or positioning constraints. The focus shifts upstream, which is the trade-off you’re accepting.
When this approach works best: accounts with strong conversion signals, clean landing-page architecture, disciplined exclusions, and sufficient volume for experiments to mature. When it fails: weak conversion definitions, messy site structure, thin negative coverage, and teams mistaking short-term CPA stability for proof of quality.
The second deadline is February 2027
September 2026 marks only phase one. Dynamic Search Ads will follow, with migration pushed to February 2027. New DSA creation will stop in January 2027, with automatic migration of remaining DSA campaigns beginning in February. This delay allows more time for preparation but also risks neglecting phase two until notices arrive in the account.
Thus, build a two-track plan: first, manage September’s ACA and campaign-level broad match migration with testing and baseline exports. Then, map every DSA dependency, including feed logic, landing page coverage, and reporting workflows, before January 2027 closes the door on new creation. If your team uses scripts or internal tools, audit those now as well, as disruptions have already been flagged for advertisers relying heavily on Google Ads API support.
Google asserts that migration should preserve existing behavior as much as possible. While this may hold true for some accounts, the landscape of Search management is changing. Expect less hand-tuning, more governance, and increased pressure on signal quality, exclusions, and measurement.
This migration timeline represents a significant shift. September 1, 2026, is merely the date on the notice; the larger change is that Google is transforming Search into a system where control shifts rather than disappears. Teams that adapt to this shift will maintain their bearings, while others may find themselves chasing symptoms throughout the quarter.
