EXPLORINEER
en

Archive — US — techcrunch

Google used AI to suspend over 39M ad accounts suspected of fraud

Google has suspended 39.2 million fraudulent ad accounts in 2024, more than triple last year's total, using AI and LLMs to enhance ad safety. Key violations include scam-related, trademark misuse, and misrepresentation. The company claims a 90% reduction in deepfake ad reports, attributed to technical updates and human oversight.

Key points

  • Google suspended 39.2 million ad accounts in 2024, up from 12 million in 2023, due to suspected fraud.
  • AI and large language models were utilized to detect and prevent ad fraud effectively.
  • Violations included ad network abuse, trademark misuse, and scams, with 5 million accounts suspended for scam violations.
  • Significant removals of 1.8 billion ads in the U.S. and thousands of ads in India were reported.
  • Google enacted over 30 policy updates and received positive results in deepfake ad report reduction.
  • A team of 100 experts analyzed and addressed ad scams, assisting in the development of effective countermeasures.
  • An appeal process with human reviews is in place to ensure fairness in account suspensions.

Why it matters

  • The growing concern over ad fraud represents a significant trend in digital advertising as companies worldwide seek to secure their advertising ecosystem.
  • The introduction of AI in ad safety reflects a broader shift towards automation and advanced technologies in business management.
  • The importance of transparency and fairness in tech-driven decision-making has gained attention in light of recent scrutiny over large tech companies.

Google's aggressive approach in tackling ad fraud through AI not only highlights its technological advancements but also emphasizes the company's commitment to maintaining a safe advertising environment while addressing fairness and transparency issues.