Google is rolling out AI content labels across five ad platforms in July 2026. Here is whether you must label your assets, the exact Asset Studio click path, and what it means for your QA process.
Google Ads AI content labels went live in July 2026, not as a hypothetical policy update. Google is rolling out an "Ads AI label setting" across five ad platforms so advertisers can disclose AI-generated or AI-edited image and video creative. Google supplies the tool. The actual obligation comes from law, and in one market it lands on you.
As of July 2026, Google is rolling out an Ads AI label setting across five ad platforms so advertisers can disclose AI-generated or AI-edited image and video creative. For most DTC accounts, labeling isn't universally mandatory yet, but it's expanding, and skipping it is a growing compliance risk in the EU, India, and New York.
It's a phased rollout, not a switch flipped overnight. Check whether the setting has appeared in your own Asset Studio before you assume anything changed for you. If it hasn't shown up yet, that's normal, not a bug.
Starting July 9, 2026, Google shipped a built-in way for advertisers to disclose AI-generated or AI-edited image and video assets, rolling out gradually across Google Ads, DV360, Campaign Manager 360, Merchant Center, and Ads Editor. Worth being precise about this, because most coverage blurs it: Google is not itself ordering every advertiser to label. Its own wording is to "Add labels directly to your creatives or use the AI label setting" (Google Ads Help). Google built the mechanism. The mandate lives in regulation.
Google framed the change around what viewers see: "We're introducing additional features, including a 'How this ad was made' panel, to help people easily understand when generative AI was used to create or alter an ad" (Keerat Sharma, Google blog, July 9, 2026).
Google auto-labels assets built with its own AI tools. Anything made with Midjourney, Runway, or a freelancer's own workflow is on you to label manually.
Two more details worth knowing. The latest version of Google Ads Editor now carries an "AI label" column that "indicates whether the labeling has been activated in your account" (ppcnewsfeed.com, July 28, 2026). And separately, Google is already watermarking its own output under the hood: it embeds "non-visible SynthID digital watermarks and C2PA (Coalition for Content Provenance and Authenticity) open-standard markup into all images and videos generated within Google Ads tools" (Google Ads Help). That part is automatic, not something you configure.
Not Google's first mid-year AI feature to land unevenly, and it won't be the last one you have to chase down account by account; see AI Max's own staggered rollout and how AI is reshaping Google Ads through 2026.
Whether you need to label an asset depends on how much AI touched it, not whether AI was involved at all. This grid is our practical judgment call, not Google adjudication: Google's own wording, "create entirely new realistic assets from simple prompts, or meaningfully enhance existing assets" (Google Ads Help), never breaks down retouching, upscaling, or background swaps.
| Asset scenario | Label it? | Why |
|---|---|---|
| Fully AI-generated product photo, no real photoshoot | Yes | Synthetic imagery replacing a real product shot |
| AI background swap behind a real product | Yes | Materially AI-edited final creative |
| AI upscale of an existing real photo | Lean yes at scale | Google treats "edited using AI" broadly |
| Light AI color correction on a real photo | Usually no | Minor, non-generative edit |
| AI-written RSA headline or description | Not required | Policy covers image and video, not text |
| Human-shot asset resized via an AI tool | No | Tool touched formatting, not content |
Where the line sits, scenario by scenario. Our practical read, not Google adjudication.
When unsure, default to labeling. Not glamorous advice, but it's the safe one: over-disclosing costs nothing, under-disclosing is a compliance gap. On the RSA question, breathe easy: AI-assisted ad copy isn't in scope, the setting governs image and video creative, not text.
You add the label from inside Asset Studio, per asset, in under a minute.
This path is confirmed by Google (Google Ads Help). One other entry point, during campaign creation, ends in "Manage AI label" then "Done" instead.
This path also works retroactively, no need to redo your whole library, just review live assets one at a time.
Pilot sighting reported by SERP Alerts (X, April 27, 2026).
The AI label setting is rolling out across five surfaces on a staggered timeline, so what you see may not match a colleague's account.
| Surface | What changes | Where you'll see it |
|---|---|---|
| Google Ads | Setting live per asset | Asset studio > Asset library |
| Google Ads Editor | New "AI label" column, live | Account view, asset grid |
| Merchant Center | Product imagery covered, live | Product data review |
| Display & Video 360 | Listed by Google, no confirmed sighting | Creative workflows, when it appears |
| Campaign Manager 360 | Listed by Google, no confirmed sighting | Creative trafficking, when it appears |
Three surfaces confirmed live, two listed but not yet sighted.
Confirmed independently: "The setting is rolling out gradually across: Google Ads. Display & Video 360. Campaign Manager 360. Merchant Center. Google Ads Editor." (Search Engine Land, July 28, 2026). If your team runs Shopping alongside programmatic display, audit each surface separately. Enabling it in Google Ads doesn't cover Merchant Center or DV360, and assuming it does is how a gap gets found later instead of now.
If more than one person touches your ad creative, in-house, freelance, or agency, this setting needs to become a checkbox in your review process.
Who owns a policy setting like this matters here, the same way conversion tracking ownership gets resolved rather than assumed.
Honestly, this is a small change for most small teams: one field on a brief, one audit pass. The anxiety around it is bigger than the actual lift.
Right now the risk is compliance exposure in regulated markets, not an automatic Google Ads penalty, but the setting alone doesn't guarantee compliance: "Use of the AI label setting in Google's advertising products doesn't guarantee compliance with specific regulations. Seek legal guidance" (Google Ads Policy Help).
Here's what most coverage skips: the EU and India route the duty mainly through platforms, not advertisers. The EU AI Act's Article 50 applies from August 2, 2026 (EU AI Act, Article 50), three days after this publishes. India's IT Amendment Rules 2026 have been in force since February 20, 2026 (Khaitan & Co). Both weigh on Google, not you.
New York is the exception, and it's the one that should have your attention. Its synthetic performer disclosure law puts the duty on "persons who produce or create an advertisement," the advertiser, not Google (Governor Hochul's office). Signed as bills S.8420-A and A.8887-B, it took effect June 9, 2026, carries civil penalties of $1,000 for a first violation and $5,000 per violation after that, and reaches any advertiser whose ads hit New York consumers regardless of where the company is headquartered (Crowell & Moring).
So: are you personally on the hook? Mostly no, except if your creative uses a synthetic human likeness reaching New York, where the answer is yes, already. This is not legal advice; confirm your specific obligations with counsel.
Mislabeling isn't a deceptive ad, either: "We continue to prohibit misleading and deceptive ads, whether created with AI or not" (Google blog). Start labeling now, while it's low stakes.
Does Google Ads automatically label AI-generated ads?
Google auto-applies the label to assets made with its own AI tools. Assets made with an external tool are yours to label manually (Google Ads Help).
Which Google ad platforms show the new AI label?
Five surfaces: Google Ads, Ads Editor, DV360, Campaign Manager 360, and Merchant Center. Only three of them, Google Ads, Merchant Center, and Ads Editor, have a confirmed live sighting so far (Search Engine Land).
Do responsive search ads need an AI label?
No. Current policy covers image and video assets, not RSA headlines and descriptions (Google Ads Policy Help).
Does labeling an asset as AI-generated hurt ad performance?
No published evidence of a penalty either way. Treat this as compliance, not a CTR question.
I don't see the AI label setting yet, is something broken?
No. The rollout window was "throughout July" and reaches accounts gradually.
Does using the AI label setting make me legally compliant?
No. Google states it alone "doesn't guarantee compliance with specific regulations" (Google Ads Policy Help). Confirm with counsel if you operate in the EU, India, or New York.
This setting is one more small policy detail that's easy to forget mid-campaign. Mira, Kampaio's creative agent, generates and edits ad creative directly, exactly the point where a disclosure like this should get handled automatically instead of sitting on nobody's checklist.
Maximus tracks policy and platform changes across every account, so the next staggered rollout doesn't turn into something you discover late. You set the direction; the agents handle the details. See pricing.
Kampaio's agents generate the creative and track the platform changes that come with it, so a new disclosure setting does not sit on nobody's checklist.
See pricingSources: Google blog, July 9, 2026; Google Ads Help, 2026; Google Ads Policy Help, 2026; Search Engine Land, July 28, 2026; ppcnewsfeed.com, July 28, 2026; EU AI Act, Article 50; Khaitan & Co, 2026; Governor Hochul's office, 2026; Crowell & Moring, 2026; SERP Alerts (X, April 27, 2026). The label-it decision grid is our practical read of Google's wording, not Google adjudication. This article is informational and does not constitute legal advice; confirm your obligations with counsel.