AI is reshaping the on-chain attack and defense landscape: the monthly frequency of on-chain attacks has increased 2.8 times in the past two years, significantly increasing the security pressure on protocols.

According to BlockBeats statistics, on June 5th, the monthly frequency of on-chain attacks has increased approximately 2.8 times since the widespread adoption of AI. Using 2020-2021 as a baseline before large-scale AI adoption, the average monthly number of on-chain attacks was approximately 3.7. However, in the two years from June 5th, 2024 to June 5th, 2026, the average monthly number of on-chain attacks was approximately 10.4, totaling 250. Looking at annual data, there were 17 on-chain attacks in 2020, 71 in 2021, and 67 in 2022. After the release of ChatGPT, the number rose to 97 in 2023, 94 in 2024, 97 in 2025, and has already reached 96 in 2026 (approximately 225 annualized). Data shows that the number of attack incidents has remained consistently higher than the earlier baseline since 2023. High-frequency attack windows are concentrated between the end of 2025 and 2026. Specifically, there were 24 attacks within the 56-day window following the release of GPT-5.2 on December 11, 2025; 21 attacks within the 42-day window following the release of GPT-5.4 on March 5, 2026; and the highest number of attacks (38) within the 35-day window following the release of GPT-5.5 on April 23, 2026. On-chain attack frequency has significantly increased after the widespread adoption of AI and has remained high in several recent model release windows, indicating that cryptographic security attacks and defenses are entering a more frequent and automated phase. With the improvement of AI-assisted auditing, vulnerability discovery, and attack script generation capabilities, the importance of protocol security, smart contract auditing, and real-time monitoring is further increasing.