Google’s Ads Safety Report gave me an opportunity to explore new data showing the enormous scale of the work required to keep the advertising ecosystem clean. This year, the struggle between technology and malicious actors has reached a new level, almost like something from a science-fiction film.
AI: Security’s New Superhero
The report’s most exciting aspect is how critical artificial intelligence has become. Using advanced AI technologies such as large language models (LLMs), Google is not only removing policy-violating ads but also stopping scammers from entering the ecosystem in the first place. Whereas older machine-learning models required large datasets, the new generation of LLMs works much more efficiently. Even with far less information, these models can quickly recognize emerging threats, identify abuse patterns and distinguish legitimate businesses from scams. Over the past year, Google made more than 50 improvements to its models. This enabled more efficient and accurate enforcement while allowing teams to focus on more complex problems. AI-supported models contributed to identifying
97% of the publisher pages on which action was taken.
Billions of Threats Blocked
Some of the report’s most striking figures concern the number of ads stopped. Here are the remarkable numbers for 2024:
- 5.1 billion harmful ads removed: These ads directed users toward misleading or harmful content
- 9.1 billion ads restricted: These had limited distribution because they contained legally or culturally sensitive material.
- More than 39.2 million advertiser accounts suspended: Most were removed before they could publish their first ad. This demonstrates Google’s proactive work against malicious actors.
These figures demonstrate the success of an approach that is proactive as well as reactive. Google expanded advertiser verification to more than 200 countries and territories, increasing transparency about who is behind the ads.
The Most Common Policy Violations
The report also shows which policies advertisers violated most often. The following were the top five policy violations acted on in 2024:
- Abusing the ad network: 793.1 million ads.
- Trademark infringement: 503.1 million ads.
- Personalized advertising: 491.3 million ads.
- Legal requirements: 280.3 million ads.
- Financial services: 193.7 million ads.
Combating Scams and Fake News
Fraud, one of the internet’s biggest problems, receives substantial attention in the report. A dedicated team was established specifically to tackle AI-generated celebrity impersonation scams. As a result, more than
700,000 advertiser accounts promoting these scams were permanently suspended. The report says complaints about this type of fraud fell by
90% following this action.To help protect election integrity, more than
8,900 new election advertisers were verified, and 10.7 million election ads from unverified accounts were removed.
Publisher Policy Compliance
Google also closely monitors the websites serving ads. In 2024, action was taken against
1.3 billion pages that violated policies. The most common categories included:
- Sexual content: 1.2 billion pages
- Dangerous or derogatory content: 24.8 million pages
- Shocking content: 13.6 million pages
This sustained monitoring helps keep the open web healthy. Google’s 2024 Ads Safety Report is more than a collection of numbers and charts. It tells the story of an ongoing effort to maintain safety in a constantly changing digital environment. It demonstrates how advanced technologies such as AI can provide protection against fraud and malicious content. A safe online experience is not an unattainable dream: it is a reality that demands continuous effort. *The figures cited are taken from Google’s published (Ad Safety Report).
Comments