AI agents have significantly reduced the cost of deanonymizing cryptocurrency holders. Researchers demonstrated that an AI agent could identify 226 out of 338 Hacker News users based on their pseudonymous posts, with costs ranging from $1 to $4 per person. The entire experiment cost less than $2000.
The method involves extracting personal details from posts, identifying candidates online, and verifying matches. While the system maintains high accuracy, it only misidentified one in ten names. In a separate test, it located a specific LinkedIn profile among 89 000 Hacker News users with a 55% success rate.
Blockchain intelligence firms are integrating AI agents into their platforms. TRM Labs launched Co-Case Agent on March 25, and Chainalysis introduced its own agents on March 31. These tools allow for tracking fund movements in minutes rather than days, serving government agencies and financial institutions.
Data leaks from companies like Trezor, which exposed nearly 81 000 customer records, and Revolut, which confirmed a breach of user data, provide raw material for these investigations. Attackers combine this information with blockchain analysis to build dossiers on targets, including their names, addresses, and estimated capital.
Physical attacks on crypto holders are rising. CertiK reported 52 confirmed wrench attacks in the first half of 2026, with 33 occurring in France. Casa security director Jameson Lopp recorded 66 such incidents between January 1 and October 14, 2026, noting that attackers often rely on data aggregated from various sources.
To mitigate risks, users are turning to mixers like Mixer.Money. These services aim to break on-chain links by fragmenting incoming funds and distributing them from different sources. However, experts emphasize that maintaining privacy requires strict operational security, as users can inadvertently re-link their assets.
