Anthropic AI Research Reveals Advanced AI Agents Pose Growing Threat to DeFi Security

New research from AI safety company Anthropic indicates that artificial intelligence agents are rapidly developing capabilities that could be used to execute real-world attacks on decentralized finance (DeFi) protocols. The study demonstrates that current AI models can autonomously identify and exploit smart contract vulnerabilities, moving beyond theoretical risks. This advancement highlights a critical and escalating security challenge for the entire DeFi ecosystem, including platforms built on Ethereum (ETH) and other blockchains. Experts warn that as AI agents become more sophisticated, the potential for automated, large-scale financial exploits increases, urging protocol developers to prioritize advanced security audits and proactive measures. [news_title_rus=Исследование Anthropic: ИИ-агенты приближаются к реальным возможностям атак на DeFi