Should You Label Your AI Ads for Google?

If you’ve seen a headline this year telling you Google now demands you declare every AI-made ad, you can relax. That’s not what happened. Google didn’t introduce a global disclosure mandate in July 2026. It did something quieter and, for most Australian businesses, more useful.

Here’s the short version. Google now lets you put an AI label directly on your image and video ads without that label breaching its own rules about text overlays and watermarks. It also gave you a setting to declare an asset as AI-created, which feeds the “How this ad was made” panel people can see in My Ad Center. Mandatory disclosure does exist, but it comes from laws in the European Union, India and New York. Australia isn’t on that list. So this is a choice, not a compliance deadline. We think you should make the choice anyway, and this post explains why.

#Key summary

#What actually changed in July 2026

Two things. First, Google started permitting text or visual labels inside image and video ad creative where that creative was generated or modified using AI. That sounds minor. It isn’t. Google’s policies have long restricted text overlays and watermarks on ad assets, so advertisers who wanted to label their AI work were stuck between a legal instinct and a policy risk. That conflict is now resolved.

Second, Google gave you a proper setting for it. You can mark an asset as created or edited with AI during campaign creation, or retrofit existing assets through Asset Studio. Google can also apply labelling automatically.

The result is a disclosure system that’s available to everyone and mandatory for almost nobody. Election advertisers are the exception worth naming. If you run election ads, you still disclose synthetic or altered content through the existing campaign settings. Don’t rely on the new AI label to cover that.

#Why the “Google is cracking down” version spread anyway

Because it’s a better headline. A policy that permits something is boring. A policy that punishes you is shareable.

There’s a second reason, and it matters more. Regulation is arriving, just not here yet. The EU, India and New York all have rules requiring disclosure on certain AI-generated or edited ad assets. Google built the labelling system to serve those markets. Commentators outside those markets saw the feature, assumed the obligation travelled with it, and wrote it up as global.

If you sell into Europe, or you run ads targeting New York, this stops being optional and you should get proper advice. For a Brisbane plumber or a Perth ecommerce store selling domestically, it doesn’t apply.

#How to apply an AI label in Google Ads

For a new campaign, select your assets when Google prompts you to review them, choose “Manage AI label”, mark the asset as created or edited with AI or decline, then confirm.

For campaigns already running, go to Tools, open Asset Studio, then Asset Library. Select the asset, mark it as AI-created or edited, and save.

It takes about a minute per asset. The bigger job is the one nobody does, which is working out which of your assets actually qualify.

#The grey area nobody wants to talk about

Google’s guidance covers using generative AI to create entirely new realistic assets, and assets edited using AI tools. What it doesn’t do is draw a crisp line through the middle of normal production work.

Some cases are obvious. A product scene you generated from a text prompt is AI-created. A photo of your actual workshop, shot on a phone, isn’t.

Then there’s everything in between. You removed a background with an AI tool. You upscaled an old logo. You had AI write three headline variants and you rewrote two of them. You used generative fill to extend an image so it fits a different aspect ratio. Reasonable people land in different places on all four.

Our working rule is simple. If a customer looking at the ad would feel misled to learn AI made it, label it. Background removal doesn’t meet that bar. A generated image of a product that doesn’t exist very much does. When you’re unsure, label it. The cost is close to zero and the downside of guessing wrong the other way isn’t.

#Why label when you don’t have to

Three reasons.

The first is that the rules are moving in one direction. No market has introduced disclosure obligations and then withdrawn them. Building the habit now, while it’s voluntary and low stakes, is cheaper than retrofitting it across every asset you own the month a rule lands.

The second is that you probably already sell to a regulated market and haven’t checked. Ecommerce is the obvious case. If you ship to the EU, or you’ve ever run a campaign that touches New York, your exposure isn’t theoretical.

The third is the one we’d actually lead with. Labelling is becoming a trust signal. Australian consumers are getting sharper at spotting generated imagery, and the reaction to being fooled is worse than the reaction to being told. An ad that says how it was made looks like a business with nothing to hide. In a market where every competitor is quietly generating creative and saying nothing, that’s a small, cheap differentiator.

#The five-minute audit worth doing today

List every tool that touched your ad creative in the last 90 days, including the ones your designer uses.

Flag anything generated rather than assisted. Generated means the tool produced something that wasn’t there. Assisted means it cleaned up something that was.

Check who has access to your ad account and whether they know this setting exists. In most small businesses the answer is that a contractor has access and nobody has ever discussed it.

Write down the rule you’ve decided to apply, in one sentence, somewhere your team can find it. Consistency matters more than where you draw the line.

Then apply labels to the assets that qualify. That’s the whole job.

#Frequently asked questions

#Do Australian businesses have to disclose AI-generated ads?

Not under Google’s advertising policies. Disclosure obligations currently come from laws in the EU, India and New York. If you only advertise to Australian customers, labelling is voluntary. If you sell internationally, get advice specific to the markets you’re in.

#Does using Google’s AI label keep me legally compliant?

No, and Google says so directly. Its guidance states that using the AI label setting doesn’t guarantee compliance with specific regulations, and advises seeking legal guidance for your local obligations. Treat the label as a tool, not a legal shield.

#Will labelling an ad as AI-made hurt its performance?

We haven’t seen evidence that it does, and the labels appear in the “How this ad was made” panel rather than across your creative in unregulated markets like Australia. If you’re worried about it, run it as a proper test rather than assuming.

#What about AI-generated ad copy, not just images?

Google’s labelling system is built around image and video assets. Text generated by AI isn’t handled the same way. That doesn’t mean copy is exempt from honesty obligations under Australian Consumer Law, which applies to what your ad claims regardless of what wrote it.

#Do I need to relabel old assets?

You can, through Asset Studio, and it’s worth doing for anything still actively serving. There’s no requirement to go back through assets you’ve retired.

#The honest bottom line

The compliance panic around AI ad labelling has been overstated, at least for Australian advertisers. Nobody is coming for your account because you generated a product shot.

But the direction is clear enough that treating labelling as inevitable is the sensible bet. Do the five-minute audit. Write down your rule. Apply the labels to the assets that qualify. Then get back to the work that actually moves the numbers, which is your offer, your landing page, and whether the ad is any good.

If you’re not sure which of your assets qualify, or you’ve inherited an ad account you’ve never audited, that’s exactly the kind of thing worth having someone look at once properly. Our Google Ads training covers practical compliance and account hygiene alongside the performance fundamentals.

This guide is written by Hedgehog, a DIY digital marketing consultancy specialising in small and medium businesses in Australia. We offer digital marketing consulting, coaching and training.