Firms Are Turning to Blockchain to Fight an Ad Fraud Problem AI Is Making Worse

Why It Matters

Immutable, blockchain‑backed proof layers give advertisers verifiable confidence as AI‑driven fraud erodes traditional measurement, unlocking a multi‑billion‑dollar protection market.

Key Takeaways

• Google removed 8.3 B ads, suspending 24.9 M accounts in 2025
• Gemini AI filters >99% policy‑violating ads before they run
• Hakuhodo‑LG pilot logged human‑verified impressions on blockchain, boosting CTR 50%
• Spindl traces clicks to on‑chain actions, delivering immutable conversion receipts
• Fraud losses could reach $172 B by 2028, creating $1.7‑$5.2 B protection pool

Pulse Analysis

AI‑generated ad fraud has exploded, forcing Google to block billions of ads and suspend millions of accounts. Gemini, Google’s next‑gen AI, ingests hundreds of billions of signals—account age, behavior, campaign patterns—to pre‑emptively filter out policy‑violating content, achieving a 99%+ catch rate. Yet the remaining ads still slip through, and generative AI tools now enable fraudsters to mass‑produce convincing fake creatives, bots, and devices at unprecedented scale, challenging legacy detection methods.

Blockchain offers a novel audit trail that can re‑establish trust. In Japan, Hakuhodo partnered with LG and World ID to run a “human‑verified ad network” where each impression is recorded on LG’s ledger only after a proof‑of‑human check, delivering a 50% lift in click‑through rates. Meanwhile, Coinbase’s acquisition of Spindl brings on‑chain attribution to the mainstream, linking a click to a wallet interaction or token purchase with an immutable receipt. By separating verified attention from verified conversion, advertisers gain granular, tamper‑proof evidence of both who saw an ad and what action followed, addressing the core weakness of probabilistic measurement.

The financial stakes are massive. Juniper Research projects global ad‑fraud losses to climb to $172 billion by 2028, suggesting a protected value pool of $1.7‑$5.2 billion if just 1‑3% of that spend migrates to verifiable proof systems. Early adoption is likely in crypto apps, independent CTV inventory, rewards campaigns, and high‑value performance ads where fraud risk outweighs integration costs. However, major platforms such as Google, Meta and Amazon may resist ceding measurement control, and regulatory scrutiny of biometric verification could slow rollout. If advertisers deem probabilistic metrics insufficient, blockchain‑based receipts could become a critical layer in the ad‑tech stack, reshaping how spend is validated and fraud is mitigated.