DeepFlowTech

Researchers at the University of California have found that some third-party large language model (LLM) routers pose security risks that could lead to cryptocurrency assets being stolen. Testing results show that some routers inject malicious code, steal credentials, and increase security vulnerabilities. The study recommends developers avoid transmitting private keys through AI agents and calls on AI companies to strengthen security measures.

Written by: Zack Pokorny
Translated by: Chopper, Foresight News
The implementation of AI agents on blockchain has not gone smoothly. Although blockchain features programmability and permissionless operation, it lacks the semantic abstraction and coordination layer suitable for intelligent agents. A research report released by the crypto research organization Galaxy points out that intelligent agents face four major structural frictions on-chain: opportunity discovery, trusted verification, data reading, and execution processes. The existing infrastructure is still designed around human interaction, making it difficult to support AI autonomous asset management and strategy execution. These are the core bottlenecks for large-scale deployment of intelligent agents on blockchain. The full report translation is below:
The application scenarios and capabilities of AI agents have begun to evolve. They are starting to autonomously perform tasks and are developed to hold and configure capital, discover trading and yield strategies. Although this experimental shift is still in its very early stages.

According to DefiLlama data, over the past 7 days, Ethereum cross-chain bridge net inflows reached $11.11B, ranking first, with Solana and Polygon at $497 million and $397 million respectively. Other platforms such as Base and Tron also experienced varying degrees of capital inflow.