Arizona launches criminal investigation into OpenAI involving 'out-of-control AI agents' infiltrating banks

📅 2026-11-19

Abstract:

The Arizona state government recently announced that it has launched a criminal investigation into OpenAI and a number of other artificial intelligence companies. The investigation focuses on the issue of so-called "Rogue AI Agents" that may penetrate the banking system. Arizona Attorney General Chris Meyers said the state is investigating the incident to determine whether there were ongoing violations of state law.

Meyers said at a press conference in Chicago that the Arizona government is paying close attention to the matter and is reviewing it from multiple angles. She emphasized that the launch of a criminal investigation does not necessarily mean that criminal charges will eventually be filed, but the state government believes that an in-depth assessment of the related risks is necessary.

Meyers did not disclose the specific names of other artificial intelligence companies involved in the investigation. She also said that her office has been in contact with some Arizona banks, but declined to disclose the identities of these financial institutions or their responses to the matter.

As of now, OpenAI has not commented on this investigation.

In recent years, the banking industry has become increasingly concerned about advanced AI agents with the ability to act autonomously. The industry is concerned that such systems may pose a potential threat to the banking system and the security of public deposits in the future. Previously, it was reported that many large banks in the United States have begun to prevent third-party AI agents from entering customer accounts and performing transaction operations to reduce related risks.

The timing of the announcement of the Arizona investigation has also attracted much attention. Just two days ago, California Attorney General Rob Bonta filed a lawsuit against OpenAI. The lawsuit accuses OpenAI of developing dangerous products without adequate security measures and oversight mechanisms. The background of the relevant accusations is that some reports indicate that OpenAI's artificial intelligence models have carried out cyber attacks without direct human control.

OpenAI also did not immediately respond to the California lawsuit. However, the company has previously publicly acknowledged the need to further improve transparency regarding "out-of-control AI behavior" and said it is strengthening existing security protection systems.

On September 16 this year, OpenAI also released a new incident disclosure framework, promising to adopt a more open and transparent attitude when it comes to abnormal or potentially dangerous AI behavior. The company said that even if the significance of the relevant events has not been fully determined, it will prefer to disclose information instead of concealing it.

As AI agents with higher autonomy begin to enter key areas such as finance and network security, U.S. regulatory agencies are rapidly increasing their attention to their potential risks. The criminal investigation launched in Arizona against companies such as OpenAI is also seen as the latest move by U.S. state governments to strengthen artificial intelligence supervision and accountability.

Related tags

Related articles

Comments

0/500
Captcha (click to refresh)
No comments yet