Google blocked 8.3 billion malicious ads in 2025, reshapes online safety

The fight against malicious online advertising has historically felt like an endless chase, with tech platforms identifying deceptive content only after it has reached users. However, the release of Google Ads Safety Report 2025 signals a shift in approach. Globally, Google blocked or removed more than 8.3 billion ads last year.

By leveraging generative AI, specifically Gemini, the platform is now neutralizing threats before most people even see them. Over 99% of these violating ads never made it to a user’s screen, intercepted by Gemini in real time.

In Nigeria, where millions rely on Google daily for news, business operations, and sourcing reliable services, this development is critical. Scammers have grown increasingly sophisticated, utilizing AI themselves to bypass traditional filters. In a challenging economic climate where individuals are particularly vulnerable to investment scams and fraudulent employment schemes, Google’s use of AI to counter AI-driven threats offers the Nigerian digital space some much-needed breathing room.

Instead of just hunting for specific keywords, the system now actually understands what an ad is trying to do. That shift represents the most significant platform safety upgrade in years.

Key Points from Google’s 2025 Ads Safety Report

The report highlights some real critical advancements. Here are five key points that show how this is making the Nigerian web safer:

1. Intercepting scams before they spread: Scam ads continue to target Nigerian users, including fake investment schemes and fraudulent job offers. In 2025, Google took down 602 million of these and suspended over 4 million accounts tied to them. The big difference now is Gemini doesn’t just scan for red-flag words. It reads the real intent behind the ad, catching those sneaky ones that used to slip past the old systems.

2. Protecting legitimate Nigerian businesses: For small businesses and entrepreneurs here who depend on Google Ads to grow, an unwarranted account suspension can severely disrupt operations overnight. Fortunately, the introduction of AI has helped reduce those mistaken suspensions by 80% in the past year. Legit local companies can keep advertising without fear of arbitrary disruption, while fraudulent actors are effectively blocked.

3. Fighting AI-powered threats with smarter AI: Bad actors are now using generative AI to churn out convincing fakes at scale. Google is countering that by feeding Gemini hundreds of billions of signals, account history, behaviour patterns, the works. The system can now flag and stop complex threats in under a second. It’s like the defence is finally moving as fast as the attackers.

4. Eradicating fake identities: Google has tightened its advertiser verification rules big time. You now need proper ID checks, including selfie verification, and any fake documents mean instant suspension. For Nigerians scrolling through ads, this extra layer means you can have more confidence that the person or company behind the promotion is who they say they are.

5. Turning user reports into fast action: Regular Nigerians flagging dodgy ads still play a huge role. The new AI tools helped safety teams handle four times more user reports than the year before and act on them much quicker. That community feedback loop is now lightning fast.

​The overarching data from the report paints a picture of rigorous enforcement. To stay ahead of emerging threats, Google executed the following last year:

• ​Suspended 24.9 million advertiser accounts
• ​Blocked or limited nearly 480 million web pages
• ​Rolled out 35 distinct policy updates

​The era of relying on simple rule-based filters is definitively over. By placing generative AI at the core of its safety infrastructure, Google has made it exponentially more expensive and difficult for digital scammers to operate in the region. For everyday Nigerians striving to build businesses, access trustworthy information, and navigate the web securely, this is a substantial victory. While the internet remains an imperfect space, it is fundamentally more secure today than it was yesterday.

Also read: The dangerous appeal of AI-as-a-Doctor in a broken healthcare system