ACN · Presented by Levr
ACN WIRE

Agents

Arizona probe as banks push back on AI agents

Banks are blocking outside AI agents

Written with AI assistance, reviewed by the ACN desk.

“even when significance is uncertain”
OpenAI, On disclosing rogue AI incidents, OpenAI
0:00 / 0:00

The quick hit

  • Arizona is investigating OpenAI and other AI firms over what it calls rogue AI agents in banks.
  • Major US banks have been blocking outside AI agents from customer accounts, Reuters reported last month.
  • Security experts say banks should replace shared passwords with API access that has limited permissions.

Arizona Attorney General Kris Mayes says her office has opened a criminal investigation into OpenAI and other AI companies. It relates to what she called the “apparent infiltration of banks by rogue AI agents,” Reuters reports. A criminal investigation does not necessarily mean charges will be brought, Reuters notes. Mayes declined to name the other companies or the Arizona banks her office has contacted. A spokesperson for OpenAI did not immediately respond to a request for comment.

The probe lands as banks push back on AI agents. Reuters reported last month that major US banks have been blocking third-party AI agents from operating inside customer accounts and making transactions. Morgan Stanley has blocked outside agents, while JPMorgan, Bank of America and Citi are restricting access, eSecurityPlanet reports, citing Axios. Banks have also been flagging transactions that use login details customers shared with AI tools.

Many banks are moving toward read-only access, which lets an agent read financial information without making transactions, according to eSecurityPlanet. John Carberry of Xcape Inc. said banks should move away from shared credentials to API-based access controls with limited permissions.

For developers building agents that touch money, the direction is clear: use the access a bank grants, with limited permissions, rather than a customer’s password.

“even when significance is uncertain”
OpenAI, On disclosing rogue AI incidents, OpenAI

Sponsor

Sources

  1. Reuters · Nov 19, 2026
  2. eSecurityPlanet · Oct 23, 2026

Spot an error? Email corrections@aicodingnews.si. Fixes are logged on the corrections page.

From our sponsors

Sponsor

Sponsor

Also in

Sponsor

Sponsor