According to the annual cryptocurrency security report released by blockchain security company PeckShield, hacking and fraud activities related to cryptocurrencies surged in 2024, with digital assets worth $3.01 billion being stolen, an increase of approximately 15% compared to $2.61 billion in 2023.
Centralized Finance (CeFi) and Decentralized Finance (DeFi) almost equally accounted for the losses in 2024, but the proportion of losses from DeFi platforms decreased. In 2022, DeFi platforms accounted for about 80% of the total losses of $3.6 billion, which decreased to 70% of $2.6 billion in 2023, and further dropped to 53.5% this year, indicating a possible improvement in the security of DeFi platforms.
PeckShield wrote on the community platform X:
[Image Source: PeckShield]
The report by PeckShield also pointed out that Ethereum was the most attacked blockchain, accounting for 34.8% of the total incidents, followed by BNB Chain (32.1%). The losses from Ethereum were also the highest, accounting for 47.3%, followed by Bitcoin, partially due to the increase in price.
In terms of protocol categories, the blockchain gaming sector suffered losses of $502 million, accounting for 22.4%, with PlayDapp vulnerabilities leading to losses of $290 million.
AI to help DeFi platforms defend against cyber attacks
Although the ratio of losses between DeFi and CeFi seems similar, the statistics from PeckShield highlighted significant incidents such as the $305 million hack of DMM Crypto, which had a significant impact on the rise of losses in CeFi. Critics argue that this indicates that DeFi platforms have not completely resolved the threat of cybercrime.
Supporters of DeFi believe that AI-driven tools can help strengthen platform security and counteract cybercriminals. Lingling Jiang, a partner at DWF Labs, believes that a new category called “DEFAI,” driven by AI, will emerge in 2025 as a next-generation DeFi protocol that integrates security into every transaction.
Luke Xie, co-founder and CEO of Satlayer, agrees and adds that AI-related solutions will provide early warnings “before potential attacks are fully exploited.” Xie explains that AI security solutions will help users avoid situations such as honey pot attacks and obvious rug pull scams. Data source.