Last updated: August 2026
On Thursday, August 20, 2026, Google announced an update to AI Max in Search ads aimed squarely at the biggest worry of professional advertisers: how do you know that letting AI steer your budgets and bids is a profitable decision before risking real money? The update adds three specific things: a single A/B test covering multiple budgets and ROI targets across several Search campaigns (rolling out in September 2026), the ability to run experiments while brand and location controls stay enabled, and tighter planning integration that shows the impact of bid or budget changes and applies them in one click. This breakdown covers what changed, for whom, and how to approach it intelligently as an advertiser.
Source: Official Google Ads Blog
What exactly changed in AI Max?
The update was written by Brandon Ervin, Director of Product Management for Search ads at Google, and it carries three practical messages:
First — one test for multiple budgets and ROI targets. Instead of building a separate experiment for each assumption, the new capability lets you test different budgets and ROI targets across multiple Search campaigns within a single A/B test. Google states the goal explicitly: "see exactly how scaling up your campaigns impacts your bottom line." The feature rolls out in September 2026 — comfortably ahead of the seasonal shopping quarter, and that timing is itself a message.
Second — experiments without sacrificing guardrails. An advertiser relying on brand controls or location targeting used to lose part of their certainty when testing AI. New capabilities in AI Max experiments let you run tests with those settings enabled, so you measure the impact of automation "without compromising those guardrails," in Google's words — removing a long-standing excuse for not testing seriously.
Third — planning applied in one click. Performance Planner now works more closely with your campaigns: it shows how changes like bidding or budget targets may impact your existing campaign performance, then lets you apply the suggested changes directly with a single click instead of re-entering them manually.
The whole update builds on a prior capability, one-click experiments inside AI Max, and the official support page remains the technical implementation reference.
Reading the timing: why September matters
The main feature landing in September is not a coincidence on the advertising calendar. September is the learning month before the biggest quarter of the year: Black Friday in November, December shopping, and year-end gift seasons across major markets. Peak-season campaigns need weeks of data accumulation before automated bidding settles into stability, and an advertiser who starts testing AI Max in September typically has a full window to read results and correct course before every minute becomes expensive. The practical meaning: if your real season starts in November, this announcement landed at the ideal planning moment — not a late one, as an August dateline might suggest.
There is also a managerial angle many teams miss: experiments in Google Ads need sufficient runtime to reach statistical meaning. A team that waits for November to try automation enters its peak with an unresolved decision — the worst possible position, either forfeiting automation efficiency or accepting its risks without data. September resolves the equation two months before the point of no return.
Why this update is strategic, not a footnote
The advertising market is passing through a double trust crisis: advertisers hear that AI lifts performance, and simultaneously hear stories of automated campaigns burning budget on queries with no purchase intent. Measurement is what separates the two camps. When a "with vs without" AI experiment becomes executable in one procedure with guardrails intact, the conversation moves from believing slogans to reading your own account numbers.
Note the wider context from our previous coverage: Google now mandates disclosure labels on AI-generated ads to protect user trust, and a new competitor emerged as ChatGPT ads began expanding across 31 European markets. The largest ad network is combining broader automation, higher transparency, and sharper measurement — and this update closes the third gap.
Source: Official Google Ads Blog
How to work with the update as an advertiser
- Plan for September now: pick two or three of your most data-stable Search campaigns and prepare a clear hypothesis (for example: does raising the ROI target while scaling the budget preserve margin?) so you are first in line when the feature reaches your account.
- Keep controls attached to the experiment: enable brand exclusions and location settings before the test starts, so the result is directly applicable to live operation afterward.
- Use Planner for scenarios, not prophecy: the tool estimates against your current performance and remains sensitive to seasonality; treat it as a scenario-comparison instrument, not a promise.
- Tie measurement to margin, not clicks: the ROI target AI Max tests is meaningless unless built on your true post-cost profit margin — which makes "what this means for you" a financial question before it is a technical one.
Quick comparison: experiments before and after this update
| Aspect | Previous AI Max experiments | August 20, 2026 update |
|---|---|---|
| Scope of one experiment | Narrower (one assumption per test) | Multiple budgets and ROI targets across campaigns in one test |
| Brand and location controls | Friction when testing | Work during the experiment without sacrifice |
| From plan to execution | Manual re-entry of changes | Performance Planner applies suggestions in one click |
| Availability | Live | Multi-campaign part rolls out September 2026 |
Honest limits before you rush
The announcement is official but brief, and precise implementation details (limits on campaign counts per experiment, eligibility conditions, rollout timing by market) usually surface in the Google Ads help pages at actual launch — check them on rollout day before relying on any specific behavior. And remember the golden rule of testing: a two-week result in a quiet season does not generalize to a peak season; give the experiment enough time and a full purchase cycle, and document the hypothesis and outcome in writing so the test becomes a knowledge asset for your team rather than a passing memory.
Frequently asked questions
What is new in Google Ads AI Max as announced on August 20, 2026?
Three things: a single A/B test for multiple budgets and ROI targets across several Search campaigns (rolling out in September), running experiments with brand and location controls enabled, and Performance Planner integration that applies suggested bid or budget changes to your campaigns in one click.
When does the multi-budget and multi-ROI testing arrive?
Google stated it is "rolling out in September" 2026 for Search ads accounts, without a specific day or market list, so the official help pages are the reference the moment it activates.
Can I test AI Max without giving up my brand protections?
Yes — that is the core of one of the updates: new capabilities in AI Max experiments let you run tests with brand and location controls enabled, so you measure automation's impact without dismantling those guardrails.
What does the new Performance Planner integration do?
It shows how changes like bidding or budget targets may impact your existing campaign performance, then applies the suggested changes directly to your campaigns with one click instead of manual entry.
Is this for small advertisers or only large enterprises?
The multi-campaign logic delivers the most value to advertisers running several Search campaigns with financially meaningful budgets, but the measurement mechanism itself benefits any advertiser who wants decisions based on their own numbers rather than impressions.
Connect your ad decisions to a skill that scales
Automated advertising raises the ceiling of what one person can manage, and the difference is made by whoever understands the logic underneath. The AI-supported courses and certificates page at Truescho brings together digital marketing and data analysis paths at accessible prices, and our ongoing coverage of this topic's siblings — from disclosure labels on AI ads to ChatGPT ads expanding across Europe — keeps you current before changes become mandatory.