The Agentic AI system has the ability to connect encrypted wallets and participate in on chain operations, which can be used to monitor investment portfolios, initiate transactions, pay service fees, and execute DeFi strategies. At present, AI intelligent agents are mostly in the stage of human approval and control, and the industry is restricting AI behavior through whitelisting of proxy wallets, limit control, multi signature mechanisms, and time permissions. AI agents face risks such as prompt injection attacks, erroneous transaction execution, abuse of permissions, and security risks in smart contract interactions, and regulatory and accountability issues remain unresolved.
