Google has already blocked over 5.5 billion bad ads thanks to AI

Google has announced the release of its 2023 Ad Safety Report to showcase its commitment to safe and transparent advertising for marketers and businesses.

The Ad Safety Report highlights Google’s enforcement of advertiser and marketer policies, along with the key trend for 2023: generative artificial intelligence.

Insights from Advertiser Policy Enforcement

The primary driver of increased ad policy enforcement has undoubtedly been generative artificial intelligence.

While traditional machine learning models are still being trained to detect and notify of ad policy violations, Google utilizes its generative AI tool, Gemini, for ad safety measures.

According to the ad safety report, over 90% of publisher page policy enforcement utilized machine learning models and LLMs (large language models) in 2023.

Another reported metric is the number of halted bad ads in 2023: 5.5 billion.

The top five policies Google enforced the most in 2023 include:

• Abusing the ad network
• Trademark
• Personalized ads
• Financial services
• Legal requirements.

Another key trend in 2023 was the rise of scams and fraud across online platforms.

Digital marketing isn’t immune to these types of scams, which is why Google took the following measures to combat these threats:

• The Limited Ads Serving policy was launched in November 2023. The policy aims to protect users by restricting the reach of lesser-known advertisers with no established track record of good ad behavior.
• The misrepresentation policy was updated in late 2023. This policy enables a quicker response from Google to suspend accounts of bad actors.
• Google blocked over 1 billion ads for violating the “abusing the ad network” policy.

November 2023 submission policy update

The policy update introduced in November 2023 aims to provide a faster response from Google regarding the suspension of accounts of bad actors. Google blocked over 1 billion advertisements for violating the “abuse of the ad network” policy.

Nearly 7 billion restricted advertisements in 2023

Another aspect of ad safety involves the restriction or limitation of ads.

According to Google’s report, they restricted 6.9 billion ads in the following content categories:

• Legal requirements
• Financial services
• Gambling and games
• Adult content
• Copyright
• Healthcare and medicine
• Alcohol

However, restricting ads does not mean they were completely blocked. The categories listed above may be shown to users but in a significantly limited capacity.

Google invests in election integrity

During each election, users are inundated with a plethora of political advertisements.

2024 will be no exception.

According to the ad safety report, Google takes the following measures to ensure voter trust in advertising content.

Identity verification process

Transparency requirements for election ads

Restrictions on how election advertisers can target their ads

Must include a “paid for” disclosure in their ads

Publisher Enforcement Insights

Another aspect of Google’s ad safety report is the enforcement of restrictions on publishers who monetize their content.

In 2023, Google took action against over 2 billion pages.

The top five areas where Google applied restrictions the most included:

• Sexual content
• Dangerous or derogatory content
• Weapons promotion and sales
• Shocking content
• Online gaming.

Conclusion

The use of generative artificial intelligence continues to increase, likely leading to a rise in advertising fraud and more restrictions in the industry.

While artificial intelligence and machine learning increasingly detect advertising discrepancies, advertisers should remain vigilant in monitoring their accounts for potential false flags. If you are a new advertiser on Google this year, it is important to note that, due to Google’s advertising safety policies, your ads may be restricted until they learn who you are as an advertiser.

You can read the full 2023 Advertising Safety Report here.

This article is available in Ukrainian.