Google's updated Ads Terms of Service took effect on 1 July. The new language lets Google use advertiser inputs across its AI features and automated tools, while making clear that advertisers remain responsible for whatever those systems generate. No action was required, which is exactly why most advertisers have not read them.
You remain obliged to review, approve or remove anything the machine generates in your account. The AI writes, you own the consequences.
Google's rewritten Ads Terms of Service took effect on 1 July and they required no action from advertisers, which is exactly why most advertisers have not read them. The short version: Google gets broader licence to use what you put into the platform, its AI gets broader authority to build and optimise on your behalf, and responsibility for the output stays entirely with you.
The updated terms expand the language around how advertiser-provided inputs can be used across Google Ads features to improve campaign performance. That includes what you type into conversational experiences, the chat-style campaign tools Google has been threading through the interface. Your prompts, assets and campaign information can feed the systems that generate and optimise ads.
The terms also confirm Google's automated features may format, select or generate targets, ads and destinations on an advertiser's behalf. Coverage of the change notes the previous terms gave advertisers more explicit opt-in and opt-out points around automation. The direction of travel is one way, with decision authority moving to the system.
The date Google's updated Ads terms took effect, expanding how advertiser inputs feed its AI while leaving liability with the advertiser
Why it matters
The gap between authority and accountability is the whole story. Google's systems can now generate ad copy, pick landing pages and adjust targeting with fewer explicit permission gates, but if an auto-generated ad misrepresents your offer, breaches industry advertising rules or sends traffic somewhere embarrassing, that sits with you. For Australian businesses in regulated categories, health, finance and legal services among them, machine-written claims are not a hypothetical risk. The ACCC does not accept the algorithm did it as a defence.
There is a competitive-intelligence angle too. Once inputs into conversational tools can be used by Google's systems broadly, the campaign strategy you describe to the chat assistant is information you have handed over, not a private working note.
Mostly this formalises the deal Smart Bidding and Performance Max already implied. The platform automates, the advertiser supervises. The terms just put it in writing while nobody was looking.
What to do about it
Audit your auto-apply recommendations settings this week. That is where Google's automation authority actually turns into changed campaigns. Switch off anything you would not approve manually.
Put a weekly review of auto-generated assets in someone's job. Search themes, generated headlines and automatically created assets in Performance Max all count as things you have legally approved by leaving them live.
Be deliberate about what goes into conversational campaign tools. Describe what the campaign needs, not your margin structure or launch calendar.
If you are in a regulated category, get compliance eyes on generated assets before scale, not after a complaint.
The terms are the honest version of the relationship now. Read them like a contract, because they are one.