Treasury warns of AI cybersecurity threat to your bank account
Most government warnings about cybersecurity arrive wrapped in bureaucratic language that makes them easy to tune out. This one from Treasury Secretary Scott Bessent was different.
Asked on Fox News' "Sunday Morning Futures" whether Americans should be worried about artificial intelligence being used to hack their bank accounts, Bessent gave a two-word answer: "You should."
That reply landed after weeks of behind-the-scenes activity that most Americans didn't see. In April, Bessent and Federal Reserve Chair Jerome Powell held an unscheduled meeting in Washington with the CEOs of major Wall Street banks, CNBC noted.
It was an emergency gathering prompted by a new AI model from Anthropic that had demonstrated the ability to identify thousands of software vulnerabilities across major operating systems and browsers.
"What we've had in the past month was a step change in the power of one large language model," Bessent said on Fox News. "But we're going to see it from the other AI companies."
The warning isn't hypothetical. And the government's response is already underway.
What triggered emergency meeting of Bessent, Powell, and Wall Street CEOs
The April gathering of Treasury, the Federal Reserve, and major bank executives wasn't a routine financial stability check-in. It was convened in direct response to concerns surrounding Anthropic's Claude Mythos.
The new AI model has demonstrated the capability to identify and exploit cybersecurity vulnerabilities across all major operating systems and browsers, according to Federal News.
Access to Mythos is currently restricted to select authorized companies, including Microsoft, Google, Amazon, and Apple, under government oversight, according to the same reporting.
But the concern driving the emergency meeting wasn't limited to one model. Bessent made clear that the issue is structural.
Related: Bessent's tax move could sabotage your next refund
"We're going to see it from the other AI companies," Bessent said, signaling that Claude Mythos is the first visible example of a broader capability shift that regulators are now racing to get ahead of.
Federal Reserve Chair Jerome Powell framed the stakes plainly at the meeting, Federal News confirms.
"Cybersecurity remains one of the most critical risks facing the banking sector, and advancements in AI are likely to intensify these challenges," Powell said, adding that the Mythos model "has raised concerns due to its capability to identify and exploit cybersecurity vulnerabilities across all major operating systems and browsers."
JPMorgan Chase CEO Jamie Dimon echoed the concern. "Cybersecurity remains one of the most critical risks facing the banking sector, and advancements in AI are likely to intensify these challenges," Dimon said.
Treasury's public-private cybersecurity initiative mobilizing financial sector
The Sunday, May 3, warning wasn't the starting point. Treasury has been building its AI cybersecurity infrastructure since February.
On Feb. 18, 2026, the U.S. Department of the Treasury confirmed the conclusion of a major public-private initiative to strengthen cybersecurity and risk management for AI in the financial services sector.
More Personal Finance:
- Fidelity has a warning for anyone who left a 401(k) at an old job
- Living trusts: what they do and who needs one
- Fidelity sounds alarm on 401(k)s, IRAs
The initiative, developed in partnership with industry and federal and state regulatory partners, is releasing a series of six resources designed to enable secure and resilient AI deployment across the U.S. financial system.
"This work demonstrates that government and industry can come together to support secure AI adoption that increases the resilience of our financial system," Bessent said in the February announcement.
PNC Chairman and CEO William Demchak, who participated in the initiative as an executive member of the AI Executive Oversight Group, described it as "an important step to address complex challenges posed by AI."
Bessent framed the policy challenge as a balancing act, encouraging private-sector AI innovation that keeps the U.S. ahead of rival nations and malicious actors, while ensuring government oversight maintains safety guardrails.
He didn't outline specific defensive actions being taken by individual banks or regulators on Sunday, but the February initiative suggests those frameworks are being built in parallel with the public warnings.
What the AI cybersecurity threat means for everyday Americans, financial system
The concern Bessent and Powell are raising isn't abstract. AI tools capable of identifying software vulnerabilities at scale represent a qualitative leap in the threat environment facing financial institutions. One that traditional cybersecurity defenses weren't designed to counter at the speed these models operate.
OpenAI CEO Sam Altman has called for comprehensive policy measures to address the broader economic and safety risks of advanced AI. He has warned that without thoughtful policies, "AI could widen inequality by compounding advantages for those already positioned to capture the upside while communities that begin with fewer resources fall further behind."
For ordinary Americans, Bessent's "you should" is a prompt, not to panic, but to pay attention. Strong, unique passwords, two-factor authentication, and vigilance around phishing attempts remain the most effective individual defenses. At the institutional level, the February public-private initiative and the April emergency meeting signal that regulators are treating AI-enabled cyber risk as a top-tier financial stability concern, not a future-state problem.
The step change in AI capability that Bessent described has arrived, and with that, the race to build defenses fast enough to match it is now officially on.
Related: Treasury Secretary delivers surprise take on economy
The Arena Media Brands, LLC THESTREET is a registered trademark of TheStreet, Inc.
This story was originally published May 4, 2026 at 11:50 AM.