Undisclosed OpenAI incident raises fresh concerns over rogue AI agents

Just weeks after hundreds of OpenAI agents went rogue and hacked into a billion-dollar company, a previously undisclosed incident has emerged in which autonomous AI agents reportedly took control of a German-language website.

A swarm of OpenAI agents allegedly took over a German-language wiki in May and used the site as a digital bulletin board to exchange information and discuss ways of circumventing restrictions, according to an exclusive Reuters report published on September 4.

A team of researchers, including Sydney Von Arx, chief executive of AI safety organisation Nightingale, and AI researcher Cormac Slade Byrd, documented the alleged AI-agent breakout in a paper shared exclusively with Reuters.

The researchers said they discovered the activity in late August while searching the internet for indications of potentially unauthorised behaviour by autonomous AI agents.

They identified nearly 15,000 edits allegedly made by AI agents on DseWiki, a German-language wiki aimed at programmers that allows users to make collaborative contributions.

The incident has intensified concerns over increasingly autonomous AI systems that could exploit security vulnerabilities, evade safeguards and coordinate with one another in unexpected ways.

Reuters reported that OpenAI administrators were aware of the German incident weeks before the report emerged. The company did not publicly disclose it while dealing with the fallout from a separate security incident in July involving the open-source AI platform Hugging Face.

OpenAI said it could not meaningfully respond to the allegations before reviewing the report and said Reuters and the paper's authors had declined its request for access to the findings.

“We are unable to meaningfully respond to claims or findings on a report that we have not had an opportunity to review,” an OpenAI spokesperson said. “Reuters and the report's authors declined our request for access. We will carefully review its contents upon publication and take any necessary next steps.”

Some OpenAI investigators sought to determine whether the German incident was evidence of a broader pattern in the behaviour of AI agents.

However, according to sources cited by the news agency, efforts to expand the investigation encountered opposition from other employees, including members of OpenAI's legal team.

OpenAI disputed that account.

“Claims that our legal team discouraged investigation of the incident are false,” a spokesperson told Reuters.

The company also said it had acted in good faith by cooperating with outside experts and disclosing relevant incidents. It added that the activity in Germany was unrelated to the separate Hugging Face incident and therefore would not have been included in a report about that episode.

By Nazrin Sadigova