JPMorgan Chase (Chase) · AI used against the institution · severity 1 of 5
Wall Street Journal columnist Joanna Stern reported on 28 April 2023 that an AI clone of her voice, made with ElevenLabs from about 90 minutes of her recorded audio, fooled the voice-biometric system on her Chase credit card's customer-service line: the system recognised the bot as her and connected the call to a representative. It was a controlled test on her own account, with no transaction or loss. A Chase spokeswoman said voice biometrics is one of several verification tools and that customers must provide additional information to complete transactions. On 4 May 2023 Senate Banking Committee chairman Sherrod Brown cited the WSJ test in letters asking JPMorgan Chase and five other firms how they secure voice authentication against AI-generated voices.
Record last reviewed Oct 21, 2026 · Markdown
Date
Institution
Country
United States
AI system
ElevenLabs voice clone · Generative AI voice clone used against a bank's voice-biometric caller verification · vendor: ElevenLabs
AI?
Described as AI by the institution or regulator
The institution, a regulator or a court called the system AI.
Category
Status
Reported
Severity
1 of 5
Impact
No customer harm; the test used the journalist's own account · No transaction was attempted or completed. The test was cited in a congressional oversight letter to the bank.
What happened?
For a column on AI voice and video clones, Stern's producer uploaded about 90 minutes of her voice from earlier videos to ElevenLabs, which produced a clone in under two minutes. She then had "AI Joanna" call her Chase credit card's customer-service line. "At the biometric step, when the automated system asked for my name and address, AI Joanna responded. Hearing my bot's voice, the system recognized it as me and immediately connected to a representative," she wrote. When a video intern tried to imitate her voice, the system asked for further verification.[1]
The demonstration did not go beyond the connection to a human representative, and no transaction was attempted. The Verge and other outlets picked up the video the same day, describing it as spoofing Chase's voice biometrics.[1][2]
On 4 May 2023 Senator Sherrod Brown, chairman of the Senate Banking, Housing, and Urban Affairs Committee, wrote to JPMorgan Chase chief executive Jamie Dimon and to five other large firms. The letter cited a February 2023 Vice demonstration and said that on 28 April "the Wall Street Journal explained how their reporter used the same voice creation service to fool her bank's credit card voice authentication system". It asked how the banks assess voice authentication against AI, whether they use customer voice data to train their tools, and how they respond to breaches, with answers due by 18 May 2023.[3]
When did it happen?
- Apr 28, 2023
The Wall Street Journal publishes Joanna Stern's column and video; her ElevenLabs voice clone passes Chase's credit-card voice-biometric check.[1][2]
- May 4, 2023
Senate Banking chairman Sherrod Brown writes to JPMorgan Chase and five other firms about AI voice clones and voice authentication.[3]
- May 18, 2023
Deadline set in the Senate letter for the banks' answers.[3]
What was the impact?
- Customers: No customer harm; the test used the journalist's own account[1][3]
- Other: No transaction was attempted or completed. The test was cited in a congressional oversight letter to the bank.[1][3]
Severity 1 of 5: Contained, no customer harm: a journalist's controlled test on her own account with no transaction or loss, though it prompted a Senate Banking Committee letter to the bank.
How did JPMorgan Chase (Chase) respond?
A Chase spokeswoman told the WSJ that the bank uses voice biometrics, along with other tools, to verify callers, that the feature is meant to let customers identify themselves quickly and securely, and that customers must provide additional information to complete transactions and other financial requests. No public Chase response to the Senate letter was found as of 2026-10-21.[1]
What was the outcome, and did regulators act?
No regulatory action followed. The episode was one of two journalist demonstrations cited in the Senate Banking Committee chairman's May 2023 letters to six firms, including JPMorgan Chase, Bank of America and Wells Fargo, asking how voice authentication is protected against AI-generated voices.[3]
Did an AI voice clone really fool Chase's voice ID?
Yes, according to the Wall Street Journal's own test in April 2023. Columnist Joanna Stern wrote that her ElevenLabs voice clone answered the automated questions and Chase's voice-biometric system recognised it as her, connecting the call to a representative. A human imitation of her voice was asked for more verification.
Was any money taken from a Chase account?
No. It was a controlled demonstration by a journalist on her own credit-card account, and no transaction was attempted. Chase said customers must provide additional information beyond voice biometrics to complete transactions and other financial requests.
What did regulators or Congress do after the WSJ voice-clone test?
On 4 May 2023 Senate Banking chairman Sherrod Brown sent letters to JPMorgan Chase and five other firms citing the WSJ and Vice demonstrations, asking how they secure voice authentication against AI and handle customer voice data. No public responses or regulatory action have been found.
When AI goes wrong at a bank, the next morning's brief says so.
when one of these regulators moves, the next morning's brief says so · six sourced stories · 7 am PT · free
plus every regulation, bank and agent page update, the morning after · leave any morning