September 30, 2026
•
min read

Google AI Max for Search Campaigns: What It Is, How to Turn It Off, and Best Practices

Young man with curly hair wearing a black shirt outdoors against green foliage background.


Alexander Perleman
, Head Of Product @ groas
Ex-Goldman Sachs and Stanford Computer Science

alex@groas.ai

LinkedIn

AI Max is Google's broadest automation layer for Search. It expands your matching, builds assets, and sends traffic to more relevant landing pages — which is great for coverage and risky for control if you run it without guardrails.

This guide answers what it is, how to turn it off, and how to run it without burning budget. The short version: use it for discovery and scale, contain it with structure and conversion signal quality, and don't pay a human team to babysit toggles weekly.

What is AI Max in Google Ads?

What is AI Max in Google Ads? It is a setting inside Search campaigns that lets Google expand beyond your keywords using broad match expansion, automatically created assets, and final URL expansion.

In practice, that means:

  • Search term matching: Google matches queries it predicts are relevant, not just queries close to your keywords
  • Text customization: Google generates headlines and descriptions based on your ads, keywords, and landing pages
  • Final URL expansion: Google can send the click to a different page on your site it thinks will convert better

Think of it as Search running more like Performance Max. You still set budgets, conversions, bidding, brand controls, and locations. Google decides much more of the matching, messaging, and landing page.

Google detailed the controls in its official Google Ads help documentation for 2025, and outlined the intent in its official announcement in 2025: less manual keyword shaping, more AI-driven matching tied to your site and creative.

For step-by-step setup and real performance data on impressions, conversion rates, and CPA, see our breakdown of AI Max setup and results.

AI Max example: what actually changes

Say you sell commercial HVAC maintenance in Dallas. Your exact keywords are around [commercial hvac maintenance dallas] and [rooftop unit repair].

With AI Max off, you only enter auctions for close variants of what you built.

With AI Max on, one AI Max example looks like this:

  • A search for "warehouse ac tune up plan dfw" matches even though you never added it
  • Google rewrites a headline to "Warehouse AC Maintenance Plans in DFW"
  • Google skips your generic Contact page and lands the user on your Maintenance Plans page

That is the upside. The downside is the same mechanism: if your conversion tracking is soft, your landing pages are thin, and you have no negatives or brand guardrails, that expansion finds irrelevant demand fast.

Google AI Max search campaigns in 2026

Google AI Max search campaigns in 2026 are no longer a limited beta. Most Search campaigns now show AI Max as an opt-in setting at campaign creation and in campaign settings.

What advertisers reported from the beta reviews and feedback in 2025 was consistent:

  • More impressions and broader query coverage, especially for services and long-tail ecommerce
  • Higher click volume with mixed intent quality unless negatives, brand inclusions/exclusions, and location intent were tight
  • Better results when accounts had strong conversion signals — qualified leads or profit-weighted purchases, not page views or raw leads
  • Worse results when accounts treated it like "set and forget" on low budget with one campaign and no structure

The pattern is familiar from our work on Search vs Performance Max hybrids: expansion without ownership cannibalizes what was working. Expansion with ownership finds net-new revenue.

Google AI Max ads: what you still control

Google AI Max ads do not remove you from the account. They change where your leverage is.

You still own:

  • Budget and portfolio vs. campaign bidding
  • Primary conversion actions and values
  • Brand inclusions and exclusions
  • Location targeting and presence settings
  • Asset quality, landing page quality, and offer clarity
  • Negative keywords and irrelevant category blocks
  • Structure: what should expand, and what must stay exact

What you give up is keyword-level precision. AI Max is intentionally keyword-light. If your strategy depends on tight ad group sculpting and match-type control, it will fight the setting. That is why keyword-centric management is failing in 2026 — Smart Bidding does not optimize at the keyword level anymore.

That is also why groas runs a different operating model. groas is a fully autonomous growth engine for paid search and organic search. Hundreds of specialized models execute bids, budgets, keywords, ads, and landing pages 168 hours a week, while a named account manager owns direction, guardrails, and result. Ad auctions change 24/7. A team that checks in once a day cannot keep up.

How to turn off AI Max in Google Ads

How to turn off AI Max in Google Ads? Turn it off at the campaign level in Settings. Google has moved the UI during rollout, but the path is the same:

  1. Open Google Ads and select the Search campaign
  2. Click Settings in the left menu
  3. Open the AI Max for Search campaigns section
  4. Toggle AI Max Off
  5. If expanded, turn off Text customization and Final URL expansion individually
  6. Click Save
  7. Check account-level brand controls and negatives still apply as you expect

Expected outcome: your campaign returns to keyword, match type, and asset control. Impression volume typically drops, irrelevant query categories stop, and CPA stabilizes within 1-2 conversion cycles.

When to turn it off:

  • You run lead gen with strict qualification and AI Max is buying informational or job-seeker queries
  • You have low conversion volume and expansion starves learning
  • You need strict message or landing page control for compliance, legal, or medical offers
  • Brand search is being rewritten or redirected away from the right page

When not to turn it off account-wide: one bad campaign does not mean AI Max is bad. Isolate it, contain it, and test it where you have budget to learn.

Best practices: run AI Max with constraints first

Do not start with "should I turn it on?" Start with budget, data, and what you can afford to learn.

1. Budget constraint

  • Under $3k-$5k/month per campaign: Keep AI Max off or to a single test campaign. You do not have enough conversions to teach expansion what a good customer is.
  • $10k+/month with 30+ qualified conversions/month: Test AI Max in one intent segment, not across everything at once.
  • Scaling accounts: Never expand budget and turn on expansion the same week. Fix signal first, then expand reach.

Increasing budget without fixing structure usually raises costs, not revenue. AI Max magnifies that.

2. Data and conversion constraint

AI Max is only as good as your primary conversion. Before you enable it:

  • Track one primary qualified action, not five micro-actions
  • Use value rules or qualified values if lead quality varies
  • Separate brand and non-brand conversions so expansion does not hide behind brand
  • Fix tracking, GA4 import, enhanced conversions, and consent gaps first

If you feed it raw leads, it will buy you more raw leads.

3. Structure constraint

  • Keep brand exact separate from non-brand expansion
  • One expanding campaign per intent cluster, not one mega-campaign
  • Add negatives proactively for jobs, DIY, free, and irrelevant use cases
  • Use brand exclusions/inclusions deliberately when running Search alongside Performance Max
  • Keep landing pages specific. Final URL expansion cannot fix a generic site.

See what AI Max does versus how groas operates differently if you are deciding between native automation and fully managed autonomous execution.

4. Team constraint

Human teams review accounts periodically. Auctions move continuously. That gap is where AI Max waste hides.

If you run in-house or with a freelancer, limit AI Max to what you can actually review weekly: search term insights, asset performance, landing page swaps, and budget pacing. If no one owns it daily, keep it contained.

groas was built for this gap. The engine blocks irrelevant keywords, avoids costly bids, uncovers cheaper high-quality traffic, and runs thousands of tests around the clock. Every action is logged with reasoning in plain language, and nothing runs outside the budgets and guardrails you set.

Why is a demand-led strategy specifically critical when using AI Max or Performance Max?

Why is a demand-led strategy specifically critical when using AI Max or Performance Max? Because both products create supply. You have to tell them what demand is worth buying.

Without a demand-led strategy, AI Max and Performance Max optimize toward any conversion they can find. That means:

  • Broad queries that look relevant but never close
  • Landing pages that get traffic because they are crawlable, not because they sell
  • Budgets flowing to easy brand or low-value wins while high-margin segments starve

A demand-led strategy flips it:

  1. Define qualified demand first: which searches, industries, products, and margins matter
  2. Build structure around that demand, not around keywords
  3. Point automation at profit-weighted signals
  4. Feed it landing pages that adapt to search intent
  5. Measure closed-won revenue, pipeline, ROAS, and CPA — not clicks or raw leads

That is the core of how groas manages paid search: outcome-oriented management toward qualified pipeline and revenue, trained on $500B+ in profitable spend, driving $1bn+ in attributable search revenue per year for 500+ businesses and agencies.

If you cannot name your best demand in one sentence, do not turn on maximum expansion.