AI Voice Cloning Hits Wall Street: Citadel, Point72 and Two Sigma Targeted in Vishing Wave

The same week Bitcoin holders were reeling from the Coldcard hack, the hosts pointed to a new report showing that AI powered attacks have jumped from crypto wallets straight into the heart of traditional finance.

THE ATTACK

Point72 Asset Management, Millennium Management, Two Sigma Investments, and Citadel were all targeted in a coordinated wave of cyberattacks in recent days, with attackers using AI-powered voice phishing to attempt to extract sensitive data from employees of the Wall Street money managers. The attackers allegedly used technology capable of listening to phone calls and then replicating a speaker's voice, tone and phrasing to fabricate convincing fake calls.

Two Sigma, which manages $75 billion in assets, told Bloomberg it successfully blocked the attempt. Spokespeople for Millennium, Point72, and Citadel declined to comment to Bloomberg.

FROM COLD WALLETS TO HEDGE FUND PHONE LINES

On the show, the hosts framed this as confirmation of a pattern they had been tracking since the Coldcard exploit surfaced: AI is lowering the cost and difficulty of large scale attacks across every kind of software, not just Bitcoin wallets. The attacks add to a string of cybersecurity incidents hitting wealth management and financial advice firms this year.

The hosts argued that this should reframe how the industry talks about artificial intelligence. Rather than treating large language models purely as a productivity story, the same underlying capability that writes code or clones a voice can be turned toward social engineering at a scale that was previously expensive and slow.

THE SAME TECHNOLOGY IS ALSO FINDING THE BUGS

The hosts noted that the frontier open source model from Chinese developer Moonshot AI, Kimi K3, has reportedly been used by Bitcoin security researchers to hunt for the kind of vulnerability that caused the Coldcard losses, something the hosts said American AI labs have been more reluctant to allow. On the show, the hosts said a company called Prem AI had stepped in to sponsor compute credits for Bitcoin teams running security audits with the model, though the specific figures cited on air could not be independently confirmed and should be treated as unverified claims from the segment rather than confirmed fact.

The hosts' takeaway was blunt: the same AI wave that is making attacks cheaper is also making audits cheaper, and whichever side of that race moves faster will define how costly this period turns out to be for both crypto and traditional finance.

This story comes from the Simply Bitcoin Live show. Watch the full episode.