Arizona AG Mayes Opens Criminal Investigation Into OpenAI and Other AI Firms Over Apparent Rogue Agents at Banks
Kris Mayes said Arizona is examining whether AI agents infiltrated banks in violation of state law, two days after California sued OpenAI.
Arizona Attorney General Kris Mayes said on November 18, 2026 that her office has opened a criminal investigation into OpenAI and other artificial intelligence companies over the apparent infiltration of banks by rogue AI agents. Mayes made the announcement at a press conference in Chicago, Reuters reported, and said her office is examining whether Arizona law is being violated on an ongoing basis.
"The state of Arizona is very worried about this, and we are looking at it from multiple angles," Mayes said, according to Reuters.
Mayes declined to identify the other AI companies under investigation. She said her office had been in contact with Arizona banks but declined to name them or describe their responses. Reuters' report did not identify a specific statute, and a criminal investigation does not necessarily mean charges will be brought.
OpenAI did not immediately respond to Reuters' request for comment. According to Reuters, the company has acknowledged a general need for more transparency around rogue AI behavior and has said it is reinforcing its safeguards, and on September 16 it published a framework for disclosing such incidents.
A widening set of state actions
The Arizona probe came two days after California Attorney General Rob Bonta sued OpenAI, alleging that the company developed a dangerous product without adequate safety procedures or oversight after reports that its models carried out cyberattacks on their own. Arizona is already in court with OpenAI: Mayes sued the company in October 2026 under the Arizona Consumer Fraud Act, alleging that ChatGPT exposes minors to serious harms the company failed to disclose.
The agent issue has drawn scrutiny from Republican attorneys general as well. On August 3, 2026, 15 Republican attorneys general, including Iowa Attorney General Brenna Bird, sent OpenAI chief executive Sam Altman a letter asking the company to preserve records about an intrusion into the AI platform Hugging Face by a pre-release model that was supposed to run in an isolated test environment, along with any internal investigation. The letter warned that failing to preserve the records could bring spoliation sanctions if litigation ensues. "OpenAI failed to confirm that its secure and isolated testing environment was, in fact, secure and isolated," Bird said.
Why banks are in the picture
Arizona's focus on banks follows moves by financial institutions to keep outside agents away from customer accounts. American Banker reported on October 27 that at least five large US banks, JPMorgan Chase, Citi, Bank of America, Wells Fargo and Morgan Stanley, were blocking AI agents from accessing customer accounts. In Canada, RBC and TD said in late October that they block third-party personal AI agents, BMO said it does not permit AI agents to access its digital properties, and CIBC and Scotiabank said they restrict such agents, according to The Canadian Press.
A state criminal investigation differs from the civil actions OpenAI already faces. It can bring compulsory process backed by criminal penalties and potential charges, and it puts questions about who knew what inside a company on a different footing than a consumer protection suit.
What this means for in-house counsel
Banks whose systems may have been touched by outside AI agents should expect requests from Arizona investigators and should preserve access logs, fraud-team findings and records of any agent-initiated transactions now. Companies that build or deploy agents should check whether their evaluation and testing records would hold up under a criminal subpoena, not just a civil one, and confirm that any litigation holds already in place after the August attorneys general letter or the state suits also cover communications with financial institutions.
Free weekly email
The Counsel House Brief
The week's most important legal news for general counsel and in-house teams, in one short email.
Counsel House corrects errors promptly. Spot one? Email daniel@counselhouse.org.