Imagine a nightclub with a very strict bouncer. He checks IDs, pats people down, and turns away anyone in a fake designer jacket. Now imagine that club has a billion doors, and new ones open every second. That’s Google’s ad business in 2026. The bouncer is real, the rules are real, and yet somehow, dodgy ads keep slipping past the rope.
Hi, Maya here. I usually spend my time explaining AI agents to you, but today’s topic sits right next door, because the reason bad ads keep showing up has a lot to do with the same automation trends we talk about all the time on this site.
What’s actually going on
Google continues to serve dodgy ads thanks to ongoing phishing scams and policy violations, despite ramping up verification and ad quality measures. Its policies are strict on paper. Counterfeit goods, for example, are flatly prohibited: ads can’t promote products that carry a trademark or logo identical to, or nearly indistinguishable from, someone else’s brand. The rules aim to curb counterfeit and dangerous products across the board.
And Google keeps tightening the screws. In 2026, it expanded its Limited Ad Serving policy to cover every Google Ads surface, from Search and Shopping to YouTube, Gmail, the Play Store, and Demand Gen. Accounts classified as “unqualified advertisers” get their reach restricted. Add a June 2026 spam update, tests of strength match labels, and expanded AI reporting, and you get a picture of a company that is genuinely trying.
So why does the problem persist?
The scammers aren’t just buying ads. They’re targeting the buyers.
In April 2026, agencies that manage Google Ads accounts were hit with phishing scams. Think about what that means. Instead of only sneaking bad ads through the front door, scammers go after the people holding the keys. Compromise an agency account, and you inherit its good reputation, its billing setup, and its access.
Google’s response was telling. “While we proactively monitor for unusual account activity to stop these incidents, advertisers must remain alert,” Ginny Marvin, Google Ads product liaison, wrote in a post on LinkedIn. Translation: the bouncer is watching, but he needs the regulars to speak up when something looks off. Advertisers are asked to stay vigilant and report suspicious activity.
Automation cuts both ways
Here’s where this connects to the AI agent world we cover on agent101.net. In 2026, AI-generated ad copy and auto-applied recommendations are standard in Google Ads. That’s automation working at massive scale, on both offense and defense. The catch, as ROI analysts have pointed out, is that following every automated recommendation blindly inflates spend without improving profit.
Sound familiar? It’s the exact lesson we keep learning about AI agents: automation is a power tool, not a babysitter. An agent that books your flights still needs you to confirm the dates. An ad system that writes and places copy automat
🕒 Published: