Google Ads' New Rules Put AI Mistakes on You, Even When Google Made Them
As of July 2026, Google Ads automatically enrolled every advertiser in AI-driven campaign automation and made you responsible for the results, even assets Google generates on your behalf. A separate July 9 rule now requires AI-generated ads to carry disclosure labels. Both changes shift real legal and brand risk onto business owners.
What Changed in the Terms of Service
The new Google Ads Terms of Service took effect July 1, 2026, and every account was bound automatically. There was no popup and no button to click. The key line authorizes Google and its affiliates to serve ads, including through automated features that format, select, or generate the targets, ads, or destinations on your behalf.
The old terms treated automation as optional features you could choose. The new terms treat it as the system you are already inside. In practice, tools like Performance Max and AI-assisted asset generation can now write headlines, build images, pick audiences, and choose landing pages using your account context, your product details, and the URLs you feed into Google's conversational tools.
Why Owners Are Now on the Hook
You are responsible for every asset Google's AI creates for you. The revised terms are explicit that advertisers must review, approve, edit, or remove all campaigns and ad assets, including the ones generated automatically. The automation does not shift liability to Google. If an AI-built headline makes a misleading claim or an auto-generated image breaks a policy, the advertiser owns that mistake.
That matters because the defense many owners assume they have does not exist. Saying the algorithm got it wrong is not a valid excuse under these terms. A machine can now create the ad, but a human still signs for it. For a business running spend without a weekly review habit, that is a gap worth closing this quarter.
The New AI Disclosure Labels
On July 9, 2026, Google started showing a How this ad was made panel inside My Ad Center that tells viewers whether an ad was created or edited with AI. The rollout is global and covers Search, YouTube, and Discover. It is the first Google ad-disclosure requirement to reach beyond political ads in years.
How the labels get applied
Ads built with Google's own AI tools are labeled automatically, because Google embeds an invisible SynthID signal in its generated content. You do not have to do anything for those.
Ads built with outside tools are different. If your team uses Canva, Adobe Firefly, or any non-Google generator, the label only appears when you self-disclose it, using a control in Asset Studio where you select an asset and choose Manage AI label. Google has said it will not verify whether you ticked that box honestly, which means the responsibility, and the reputational risk of getting it wrong, sits with you.
The deepfake line
One rule is not optional. AI-generated depictions of real, identifiable people are banned. If any of your creative includes a synthetic version of a real person, remove it now rather than waiting for a review.
What to Do This Quarter
Turn on a human review step
Before anything spends, have one person check every auto-generated headline, description, image, and landing page match. This is the single cheapest protection against the new liability, and it takes minutes per campaign.
Build a creative origin log
Keep a simple record of which tool made or edited each asset, whether it was AI-generated or only AI-assisted, and who approved it. When a question comes up later, that log is your evidence, not a guess.
Set your AI labels honestly
Go into Asset Studio and label any assets your team made with outside AI tools. The short-term discomfort of a label is far cheaper than the long-term cost of being caught hiding one.
Do not treat the checkbox as legal cover
Google's own help documentation says the AI label setting does not guarantee compliance with specific regulations. Advertising, disclosure, and consumer-protection laws still apply on top of Google's rules, so treat the label as a starting point, not a finish line.
The Bigger Picture
The pattern here is bigger than one policy update. Platforms are handing owners more automation and, at the same time, more accountability for what that automation does. The businesses that stay safe will be the ones that pair the speed of AI with a clear human checkpoint, a paper trail, and honest disclosure. That is not a reason to avoid automation. It is a reason to run it on purpose.
At TMC Marketing, we manage paid media for growing companies with exactly these guardrails built in, so owners get the efficiency of automation without carrying the risk alone. If you want your Google Ads reviewed against the new rules before your next spend, Schedule Discovery Call.
Frequently Asked Questions
Do I have to accept the new Google Ads terms?
No. The revised Terms of Service took effect July 1, 2026, and every account was bound automatically with no prompt or acceptance click. If you are still running ads, the new terms already apply to you.
Am I really responsible for ads that Google's AI created?
Yes. The terms state that advertisers must review, approve, edit, or remove all campaigns and assets, including automatically generated ones. Liability does not shift to Google, so a human on your side still owns the final result.
Which of my ads will get an AI label?
Ads built with Google's own AI tools are labeled automatically through an embedded SynthID signal. Ads made with outside tools like Canva or Firefly are only labeled if you self-disclose in Asset Studio, and Google will not verify your answer.
What happens if I do not disclose an AI-made asset?
Google will not check the box for you, but the reputational and legal risk stays with you. Consumer-protection and advertising rules still apply, and getting caught hiding AI use can cost more than the label ever would.
Does the AI label keep me compliant with the law?
No. Google's help documentation states plainly that the label setting does not guarantee compliance with specific regulations. You still need your own review and, where stakes are high, legal input.