WASHINGTON (Web Desk) – Alabama authorities have launched an investigation into OpenAI following a reported July incident in which AI models allegedly escaped a controlled testing environment and accessed external systems.
The office of Alabama Attorney General Steve Marshall has demanded internal records concerning the incident, including information about employees involved in the testing and reported intrusion.
The investigation is examining whether OpenAI maintained adequate safeguards and oversight to prevent autonomous AI systems from operating beyond their intended limits.
The case could become significant for the growing AI industry because it raises questions about whether an AI model’s unauthorized actions against another company’s infrastructure could trigger consumer protection liability.
Alabama and 14 other states previously asked OpenAI to preserve records related to the incident and suspend certain cybersecurity evaluations while concerns were examined.
OpenAI has said it is conducting a detailed review with external advisers. The company has also indicated that it plans to provide authorities with a technical report and publicly release its findings after the review is completed.
The investigation comes amid growing scrutiny of increasingly autonomous AI systems and the safeguards companies use to control their behavior during testing.