TL;DR:

Jason Gewirtz received a fraudulent call from a supposed security agent claiming suspicious activity in Germany.
The attacker used real personal data and emails with official logos to create a false sense of urgency.
Artificial intelligence and direct contact with the platform allowed the fraud to be identified before funds were lost.

A new incident puts security in the crypto ecosystem under scrutiny. Recently, Jason Gewirtz recounted how he was nearly a victim of a hack on his Coinbase account through a highly sophisticated phone call. The attacker, who identified himself as “Brian Miller” from the security team, attempted to deceive the executive with detailed personal information and real-time unauthorized transfer alerts.

Despite the psychological pressure applied by the attacker, Gewirtz noticed key inconsistencies, such as suspicious email senders that did not match official domains. The scammer persisted and tried to prevent the victim from changing his login credentials, claiming this would freeze his funds, while attempting to guide him toward setting up a supposed “hardware wallet” controlled by the criminals.

The Role of Artificial Intelligence and Institutional Response

To verify the legitimacy of the call, Jason Gewirtz turned to the AI chatbot Claude, which immediately confirmed it was a phishing campaign. Furthermore, after contacting internal sources at the platform, it was confirmed that the company never makes direct phone calls to request fund movements or to protect accounts from external attacks.

This incident marks a worrying trend: the use of voice modulation tools by organized groups that recruit young people to execute precise technical scripts. Consequently, asset recovery firms report a 1,400% increase in impersonation scams, where attackers exploit urgency to bypass standard security protocols.

In summary, Coinbase reiterates that any instruction to move cryptocurrencies to a “secure wallet” should be considered an absolute fraud. Given the advancement of social engineering supported by AI, constant vigilance and the use of external verification tools are now the first line of defense to protect users’ digital property.

---

Filed under: News - @ January 21, 2026 4:24 am