Arizona opens criminal probe of OpenAI over rogue AI agents
Arizona AG Kris Mayes opened a criminal investigation into OpenAI and other AI firms over rogue AI agents in banks. What Arizona businesses should do with agents now.
Arizona has opened a criminal investigation into OpenAI and other AI companies over what Attorney General Kris Mayes called the "apparent infiltration of banks by rogue AI agents." Mayes announced it on November 18 at a press conference in Chicago, Reuters reported. We build software in Phoenix, so this one is on our home turf. If your business lets an AI agent near a bank account, read on.
What actually happened
Per Reuters:
- The probe: Mayes said her office is examining whether Arizona law is being broken on an ongoing basis. A criminal investigation does not mean charges will follow.
- Who: OpenAI is the only company named. Mayes declined to name the others.
- Banks: her office has been in contact with Arizona banks. She declined to name them or share their responses.
- No incident disclosed: the report describes no specific bank incident behind the probe.
- Her words: "The state of Arizona is very worried about this, and we are looking at it from multiple angles."
- OpenAI: a spokesperson did not immediately respond to Reuters.
It came two days after California Attorney General Rob Bonta sued OpenAI, alleging it built a dangerous product without adequate safety oversight after reports its models ran cyberattacks on their own. Reuters also reported last month that major US banks have been blocking third-party AI agents inside customer accounts.
Why it matters for your business
The facts here are thin. The direction is not. Regulators, banks and now a criminal prosecutor are treating AI agents that touch money as a risk. Your bank's terms may already put the loss on you.
What we tell Arizona clients:
- No bank logins in AI tools. Not in a browser agent, not in a chat window, not in a shared password note an agent can read.
- Read-only by default. Agents get transaction data through your accounting software's bank feed, not direct account access.
- A human approves money movement. The agent drafts the payment. A person releases it.
- Log every action. If a regulator or your bank asks what the agent did, you need the answer in minutes, not a guess.
- Ask your bank what its policy on AI agents is, in writing, before you automate anything.
Key takeaways
- Arizona AG Kris Mayes opened a criminal investigation into OpenAI and unnamed AI firms over rogue agents in banks
- No banks, other companies or specific incidents have been named
- A criminal investigation does not mean charges will be brought
- It follows California's lawsuit against OpenAI and banks blocking third-party agents
- Keep agents read-only on financial accounts and put a human on every payment
Running AI agents near your books? We're in Phoenix. We build agent workflows with read-only access, human approval on money movement, and a full action log. Talk to us or see how we build.
Sources: Reuters via The Star, Reuters via Yahoo Finance.
- #ai-agents
- #arizona
- #openai
- #banking
- #ai-regulation
Tommy Rush — Founder, Rush Commerce
Operator turned builder. 15+ years running operations — now shipping the systems businesses run on. More
Get The Rush Report weekly — one email, zero fluff.
Keep reading
Together AI prompt leak: an AI coding agent committed the key
Together AI says an AI coding agent pushed a credential to GitHub, exposing stored prompt samples from June and July. ZDR is now default for new accounts only.
Read itSuperhuman Go AI assistant: proactive agents, Docs on Dec 1
Superhuman Go becomes a proactive AI assistant across mail, docs and 100+ tools, with triggered Docs agents in beta Dec 1. Scope what it can see first.
Read it