The AI agent within Google Ads is getting new safety features after it revealed millions of publishers violated its policies last year.
Google is introducing three new agentic safety features in Ads Advisor, aiding the transition from a recommendation system to enforcement automation across the platform.
The tech giant is shifting into autonomous AI enforcement for advertisers with these new tools; agentic safety monitoring, automated account protection, and policy violation prevention.
These will help advertisers with flagging more complex policy violations without needing human intervention, as well as strengthening account security, and gaining instant approval for certifications, bypassing weeks of manual paperwork.
The move comes after Google released its Ad Safety report, which revealed the tech giant has removed or blocked more than 270 millions gambling ads in 2025 and handed out almost 10 million publisher violations.
“Ads Advisor will soon turn weeks of paperwork into instant approvals by proactively determining if your business needs a certificate based on your industry and country,” explained Priya Baliga, senior engineering director of ads platforms at Google.
“Then it will either grant it automatically or, if more details are needed, help you submit your application in one click. As always, Ads Advisor will ask for approval before taking action in your account.”
Platforms have traditionally required explicit human approval for suggestions and policy decisions, but the new autonomous agents remove the need for manual review.
For advertisers, this means small and medium businesses without ad operations teams can spend less time and resources on manual approvals. This is framed as a way for Google to improve efficiency and protections for advertisers, but it represents a broader migration towards automation at scale.
Advertisers get a choice. Automation brings cheaper, quicker, and more efficient advertising flow, but without a human in the loop, transparency, accountability, and appeals become more difficult.
It won’t come as much of a surprise that Google is pushing for further automation, especially with the scale of its advertising arm. Other huge advertising competitors like Amazon and Meta are likely to follow suit given the pressure that advertising platforms face to help protect brands, with huge quantities of advertisers using the services.
Agentic AI is becoming more prominent outside of the tech world, with retail, compliance teams, and now advertising agencies taking advantage of the speed and volume of tasks that can be undertaken.
The litmus test for this role out will be the adoption from bigger advertisers. Enforcement calls are a sensitive balance, and trust in AI to make unsupervised decisions is not guaranteed. Larger advertisers, with well-established brands to protect won’t remove human approval lightly.
