Patrick Steven Yaroch, a high-ranking special agent within the United States Federal Bureau of Investigation (FBI), has been arrested on charges of stealing $1 million from a cryptocurrency wallet belonging to a foreign national. A federal investigation revealed that the agent did not use any password-cracking methods to transfer the funds; instead, he memorized recovery phrases from internal systems to execute the transaction, subsequently using ChatGPT to plot an international escape.
A Million-Dollar Heist via a Simple Method
Serving in critical national security roles at the FBI's Boston field office between 2017 and 2025, Yaroch used his high-level security clearances to access a foreign individual's crypto wallets during an investigation in November 2024. By reading and memorizing the recovery phrases stored in the bureau's internal systems, the agent transferred a total of $1 million in about 10 to 12 transactions into a wallet he created, without needing to breach any encryption or security protocols. After depositing the stolen funds into the Suilend protocol on the Sui blockchain, the agent contemplated confessing after some time, but was captured in a raid conducted by federal teams.
Details of the ChatGPT-Assisted Escape Plan
Deepening their investigation, prosecutors accessed the artificial intelligence chat history on Yaroch's mobile phone. Contrary to the agent's defense that he committed the crime on impulse, it turned out he had used ChatGPT as a comprehensive escape and investment tool:
- May 28: Queried how to invest the $1 million to maximize returns.
- June 4: Researched how to leave the U.S. with the stolen funds and obtain residency or citizenship in a European Union country.
- June 17: Asked whether a visa was required when transiting through Turkey.
- June 26: Had an email draft written to an executive for a business opportunity in Greece.
Acting on these prompts, the agent granted power of attorney to lawyers in Portugal and traveled without his agency's knowledge, but his plan was upended by digital footprints.
Industry Implications and Security Vulnerabilities
This incident lays bare the massive threat that insider threats—posed by authorized personnel working in high-security institutions—represent when digital assets are involved. Despite advanced firewalls or encryption layers, the exploitation of the human factor and manual security vulnerabilities such as seed phrases demonstrates that new oversight mechanisms are imperative in institutional security protocols.
Frequently Asked Questions
How was a law enforcement officer able to access the wallets without cracking passwords?
No hacking or password-cracking methods were used in the incident. During an official investigation, the agent used his high-level authorization to view and memorize the seed phrases found in internal systems, carrying out the transfer with this information.
How was the AI chat history used as evidence in the legal process?
Although the suspect explained his crime as "sudden loss of control and psychological pressure," prosecutors obtained concrete evidence proving the plot was premeditated by uncovering weeks-long search histories on ChatGPT regarding residency, visas, and fund management.
*This news report is based on data published by Webtekno — Artificial Intelligence.
💬 Comments
No comments yet. Be the first!
You must be logged in to comment.
🔑 Log In