Google Suspended Over 39 Million Fraudulent Ad Accounts Using AI Tools

Google suspended over 39 million fraudulent advertiser accounts in 2024 using AI, blocking billions of scam ads before they reached the public.

Google implemented its largest ad fraud crackdown to date by halting 39.2 million advertiser accounts during 2024 — more than triple the previous year’s total, as first reported by TechCrunch. The company attributed this improvement in early detection to its use of AI and large language models (LLMs).

The AI models help to identify abnormal activities including fake business details and payment fraud. The majority of these accounts were halted before their ads could launch.

“While these AI models are very, very important to us and have delivered a series of impressive improvements, we still have humans involved throughout the process,” said Alex Rodriguez, General Manager for Ads Safety at Google, during a media roundtable, as reported by TechCrunch.

Rodriguez stated that more than 100 specialists from Google’s Ads Safety team, Trust and Safety division, and DeepMind worked together to analyze threats, including deepfake ads that impersonated celebrities, as reported by TechCrunch. Google suspended more than 700,000 ad accounts related to these scams, and deepfake-related complaints dropped by 90%.

TechCrunch reports that the U.S. saw the removal of 39.2 million ad accounts, along with 1.8 billion advertisements. In India, 2.9 million accounts were suspended, and nearly 250 million ads were taken down — many of which promoted gambling schemes, scams, or fake brand impersonations.

According to Mashable, Google reported that its platform eliminated 415 million scam-related ads and blocked a total of 5.1 billion ads across all regions in 2024. Many of these ads violated policies related to sensitive topics, including personal hardships, religious beliefs, and health claims.

The crackdown comes amid broader concerns over online ad security. Microsoft recently warned of a massive malvertising campaign that infected nearly one million devices. Cybercriminals used malicious ads to redirect users to GitHub repositories hosting malware, which stole data and enabled remote access.

This follows a separate Android threat: researchers recently discovered a campaign distributing SpyNote malware via fake Google Play Store pages.

These deceptive sites trick users into downloading infected apps disguised as popular software. Once installed, the malware enables attackers to access microphones, GPS, contacts, and even remotely control the device.

The large-scale suspension of accounts often raises concerns about fairness. Rodriguez acknowledged this, noting that Google has updated its policies and communication tools to improve clarity for advertisers.

“Oftentimes, some of our message wasn’t as clear and transparent about specifics […] It’s been a big focus for the team as part of 2024 and into 2025,” he said, as reported by TechCrunch.