Distributed ledger technologies (DLT) are evolving rapidly, with applications in international payments (stablecoins) and decentralized finance (DeFi) and social networks. Simultaneously, artificial intelligence (AI) systems based on Large Language Models (LLMs) are enabling advanced autonomous agents that are much more powerful than traditional bots. While bot activity in social media has been studied previously, the interaction of autonomous agents with DLT protocols is a newer field. The permissionless nature of DLTs allows these agents to engage in online economic activity, bypassing traditional financial restrictions. Already a majority of stablecoin transactions are executed by bots, and this trend is expected to continue. Moreover, decentralized social media platforms like Farcaster allow these agents to integrate into online social networks.This project aims to analyze the emerging ecosystem of autonomous agents in DLT technologies. It focuses on establishing data collection infrastructure, developing analytical methods, and characterizing agent activity across DLT protocols. Macro-level analysis will involve structural properties of agent interaction, while micro-level studies will assess individual agent behavior. Machine learning, including supervised classification and LLM-based reasoning, will help predict features such as potential malicious activity. Additionally, the team will deploy an autonomous agent on a DLT platform to study its behavior and impact.The research group at the Rudjer Boskovic Institute includes Dr. Matija Piškorec as principal investigator, Dr. Damir Korenčić as project associate, a postdoctoral researcher, and a PhD student. The project will be conducted in collaboration with the Blockchain and Distributed Ledger Technologies (BDLT) group at the University of Zurich, led by Prof. Claudio Tessone, with whom Dr. Piškorec has an ongoing collaboration.