Google's New 'AI' Label on Search Ads Will Quietly Drift Your YoY Benchmarks

Google's New 'AI' Label on Search Ads Will Quietly Drift Your YoY Benchmarks
Brodie Clark spotted two variants of the 'AI' badge on mobile search ads. The click goes nowhere, which is the part worth pricing in.

Google is testing an "AI" label on some mobile search ads, spotted April 26, 2026 by SEO consultant Brodie Clark. Two variants are running: a rounded pill and an "AI" text with an info icon. Clicking either does nothing. Paid teams should start segmenting AI-surface impressions now, because once a chunk of inventory sits behind the badge, year-over-year search benchmarks will drift without showing up in any standard report.

What the test actually looks like

The label appears on individual sponsored results in mobile search, not at the top of the SERP and not on every ad. Clark posted screenshots showing two design treatments side by side: one with a small "AI" pill on a soft rounded background, and one with the letters "AI" followed by a tiny info icon. Tap either and nothing happens. No tooltip, no definition, no link out to a help page explaining what "AI" means in this context.

That last detail is the part worth sitting with. Most Google badge tests, even ones in early beta, click through to a brief explainer. Sponsored, About this result, the redesigned ad-disclosure card from October. This one is a dead end, which usually means one of two things. Either the click handler has not been wired up yet, or the label is a regulatory placeholder that Google legal wants on screen before they have to decide what it actually discloses.

The original report ran on Search Engine Roundtable, and as of writing there is no comment from Google. Both variants are showing concurrently, which usually points at a multivariate test rather than a phased rollout.

Why Google is preparing the badge surface now

This isn't an isolated UI tweak. It lands inside a six-month string of Google moves that all expand AI inventory and AI-driven copy generation.

In November 2025, ads started surfacing inside AI Mode with a "sponsored" label at the bottom of the response, as Search Engine Land covered when the test scaled. In February 2026, AI Max for Search text guidelines opened to all advertisers worldwide. In April 2026, Google announced it would retire Dynamic Search Ads in favor of AI Max, with Phase 2 auto-upgrading legacy campaigns in September.

Add the EU AI Act content-transparency obligations now in force, and the regulatory pressure to put a disclosure surface on AI-touched ad inventory looks unavoidable. My read: Google is building the badge first so they can decide later what it discloses, without ripping the UI back in. The fact that the click does nothing right now would be embarrassing if anyone outside SEO Twitter cared, and so far they don't.

The measurement leak nobody is pricing in

Here is the part that actually matters for working media buyers, and the reason this is worth more than 200 words on a Monday morning.

When AI-labeled inventory expands, your performance reports do not get a new dimension. There is no "AI placement" filter in Google Ads. Your impressions, clicks, CPC, and conversion volume continue to roll up into the same Search campaign rows you have been benchmarking against for years.

That sounds harmless until you ask the obvious question: do AI-labeled ads convert at the same rate as non-labeled ads? They almost certainly do not. From what I have seen across other badge changes (the sponsored relabel in 2023, AI Overviews citations, AI Mode sponsored slots), any disclosure surface that changes user perception tends to move CTR by 5-15% in one direction or the other. Sometimes both, depending on query intent.

So if 20% of your mobile search ad impressions end up sitting behind an "AI" pill in Q3, and that subset converts 10% lower than your non-labeled inventory, your blended CVR drops by roughly 2 points. Smart Bidding sees the lower conversion rate as a signal change and reweights spend toward queries that still convert, which over a few weeks tightens your impression share on the AI-labeled inventory and quietly redistributes budget toward whatever Google treats as the non-AI bucket. You will not see any of this in a report. The line that moved is your year-over-year benchmark, and it moved for a reason that has nothing to do with your creative, your bidding strategy, or your audience.

We covered the sister version of this problem in The AI Overviews CTR Rebound Tracks Google's Impressions Bug, Not a SERP Shift. Same pattern, different surface. AI surfaces leak into your reports without leaving a fingerprint.

A 30-minute segmentation move worth doing this week

I would take 30 minutes before this label gets out of beta and do three things. None of them require new tooling.

1. Snapshot a mobile-only baseline. Pull a mobile-segment view in Google Ads for the last 14 days. CTR, average CPC, conversion rate, by campaign. Save it. This is your pre-label baseline. When the AI badge starts showing on more than 5% of mobile inventory (probably Q3, based on the November-to-April cadence of the AI Mode ad rollout), you will want this snapshot to compare against.

2. Stamp your final URLs with a placement parameter. Add something like ?placement_check={device}_{network}. Google does not currently surface AI placement as a dimension, but if they ever add it (and I would bet they do, once EU enforcement bites), you will have raw click logs in your analytics pivoting on the same parameter set you have been carrying for months. Cheap option value.

3. Set a calendar nudge for early August. By then the badge will either be everywhere, gone, or paired with an actual click-through explainer. Whichever way it goes, you will want to compare the August mobile baseline to the April one, not run a 12-month YoY against a quarter that didn't have AI labels at all.

And to be fair, most accounts will not see a meaningful blended-CVR move from this in the next month. The label is on a small share of inventory, and Smart Bidding tends to absorb the noise before it shows up in a weekly review. The reason to do the work now is that your reports get more honest the more cleanly you draw the pre-AI line. The cost is half an hour. The cost of not doing it is a year of slightly wrong YoY narratives in QBRs, with no way to back into the cause.

The question Google did not answer in beta

If the AI label is meant for users, the click should disclose something. If it is meant for advertisers, it should at least surface in placement reports. Right now it does neither, which probably means the badge exists for a third reason Google is not saying out loud yet.

The most plausible read is regulatory. Once you are required to flag AI-touched content in a given jurisdiction, having the disclosure surface already shipped in production is much easier than rolling out new UI under a deadline. So the click goes nowhere on purpose, until legal decides what it should say.

In the meantime, the label is on screen. Some percentage of users see it and act on it. And your campaign reports keep summing up impressions like nothing changed. I would rather have the segmentation snapshot in the bank before the gap shows up in a quarterly review.

Notice Me Senpai Editorial