Google’s 2024 Ads Safety Report reveals a sharp rise in enforcement efforts, powered by AI and large language models (LLMs).
The company focused on earlier fraud detection and stronger policy enforcement at scale.
Key Stats:
– 5.1 billion ads removed
– 9.1 billion ads restricted
– 39.2 million advertiser accounts suspended (↑ 200% YoY)
– 1.3 billion publisher pages enforced against
– 415 million scam-related ads blocked/removed
– 30+ policy updates rolled out
AI’s role:
50+ AI model upgrades enabled faster, smarter detection of fake payment info, impersonation, and other red flags.
97% of publisher violations were caught using AI.
A dedicated team of 100+ tackled impersonation scams, suspending 700,000+ scam accounts and reducing incidents by 90%.
Election-year focus:
With elections in 50+ countries, Google tightened political ad rules.
10.7 million unverified election ads removed.
AI-labeling became mandatory for election-related AI content.
Takeaway: Google is using AI to both detect scams faster and enforce policies more strictly.
But concerns remain about transparency and policy clarity for legitimate advertisers.
