Arizona has opened a criminal investigation into OpenAI and other artificial intelligence companies over concerns that AI agents may have entered banking systems without proper control. Attorney General Kris Mayes said her office was examining whether state laws had been broken. The investigation follows growing concern about the ability of advanced AI tools to carry out tasks with limited human direction. Officials are looking at whether these systems could create risks for banks and their customers.
Mayes said the state was taking the issue seriously and examining it from several angles. Her office has contacted banks in Arizona but has not named the institutions involved. She also declined to identify all the companies under investigation. OpenAI did not immediately respond to a request for comment. The investigation does not mean that charges will necessarily follow, but it shows that state authorities are examining the possible legal impact of AI agents.
AI agents can perform tasks that go beyond answering questions. Depending on how they are designed and connected to other systems, they may be able to interact with websites, software and online accounts. This has raised concerns about whether an agent could access private financial information or attempt transactions without proper permission. Banks have become increasingly cautious about allowing outside AI tools to operate within customer accounts.
The Arizona inquiry comes as regulators across the United States examine the risks created by fast-moving AI technology. California has also taken legal action involving OpenAI over claims about product safety and oversight. The wider debate is now moving beyond how AI systems generate information and toward what they can do when connected to important services. The Arizona investigation could help define how existing laws apply when automated systems interact with financial institutions and personal accounts.

