Two platforms, one name, opposite defaults
September is the first full month where Google and Microsoft are both running something called AI Max for Search. They share the same name. They share the same three core features — search term matching, text customization, and final URL expansion. But if you're managing accounts on both platforms, treating them as the same product will cost you.
Google's AI Max is mandatory. The migration runs September 1 through 30. If your campaigns used automatically created assets or campaign-level broad match, they've already been converted — no opt-out, no rollback, no notification beyond the account change history. What changed in your account on September 1 has been covered extensively.
Microsoft's version went globally available on August 19, following an April announcement and a May open pilot. The feature set is identical on paper. The delivery is not: every feature in Microsoft AI Max is individually opt-in, and you choose which combinations to enable and when.
What each feature does
All three features work the same way conceptually across both platforms.
Search term matching uses your ads, landing pages, and user intent signals to serve your ads on queries beyond your explicit keyword list. This is the feature with the biggest reach impact, and the one that needs the closest watch on search term reports when you first turn it on.
Text customization generates additional headline and description variations from your existing creative and website content. Quality varies by account and category. When it works, you get genuine variations. When it doesn't, you get generic filler that doesn't match your landing page. How often Microsoft's model is serving AI-generated text versus your original copy is something you'll need to track in reporting.
Final URL expansion routes users to the most intent-relevant page on your site rather than the destination URL you set on the ad. For accounts with well-structured landing pages mapped to specific product categories, this is often a net positive. For accounts where landing page architecture was built without this in mind, it introduces variability you probably don't want.
The auto-enabled exception
If you were running Predictive Matching or autogenerated text assets before Microsoft's global launch, those settings were automatically converted to their AI Max equivalents. The changeover happened without a campaign flag or a notification in your account.
If you've inherited accounts, or haven't done a settings audit in the last few months, check this before assuming you're on a baseline configuration. The features may already be running.
Why the opt-in philosophy helps you right now
Google's forced migration is mostly damage control at this point — verifying brand exclusions, auditing search term reports for off-target queries, checking that URL expansion isn't routing users to irrelevant pages. You're working backwards from what changed.
Microsoft's approach lets you run a controlled test. Enable one feature at a time, establish a performance baseline, and compare against the control state. That's how you determine whether AI Max is helping your specific account — not by accepting aggregate claims as a proxy for your campaigns.
The timing also matters. Q4 starts in six weeks. Any AI feature you enable needs a learning period before peak traffic. Turning on text customization or URL expansion in mid-November — when CPCs are elevated and any campaign change carries real cost risk — is not when you want the algorithm learning what your converting traffic looks like.
If you want Microsoft AI Max in good working order for Q4, start the test in September.
What the identical feature list tells you about the industry
Google and Microsoft arriving at the same three features and the same product name isn't an accident. Search term matching, text customization, and URL expansion are the three capabilities defining AI-augmented search advertising right now. Both platforms built to the same specification.
The differentiation isn't in what the features do. It's in the data each platform brings to make them work, and in the defaults each platform chose. Google has more search volume and training data. Microsoft has LinkedIn professional identity signals and a query profile skewed toward enterprise and research intent.
Whether those data differences produce better results from the same feature set depends on your category, your audience, and your account history. That's an empirical question your own account data answers — platform benchmarks won't tell you.
Four things to do this week
Open Microsoft Ads and check whether Predictive Matching or autogenerated text assets were running before August 19. If they were, AI Max is partly live already — start there.
Pull search term reports to check for unexpected query matches if search term matching was auto-converted as part of that transition.
Decide which of the three AI Max features you want to test before Q4. Enable them now, not in October. The learning period needs to run while traffic is manageable, not while you're in peak spend mode.
On the Google side: if you haven't verified brand exclusions and reviewed search term report changes since September 1, do that before anything else.
If you're not sure whether your conversion tracking and attribution setup can actually support what these platform AI tools need — offline conversion imports, proper lookback windows, signal quality — the free audit at /audit surfaces the gaps in about three minutes.
Google's AI Max is live in your account whether you chose it or not. Microsoft's is waiting for your decision.
Sources: Search Engine Land, Search Engine Journal, Microsoft Advertising Blog, August–September 2026
