OpenAI bots took control of German website, found ways to bypass restrictions, created backups when blocked | Here's how

OpenAI’s AI agents recently did something that cannot be simpy dismissed as ‘going rogue’. The AI agents cheated on tasks, communicated with each other, evaded detections, and finally tried to create backup pages when attempts were made to shut them down.

How the bots took control of the websites

The episode, which began in May on a German-language wiki called DseWiki, involved more than 15,000 edits, according to new research shared with Reuters. The site meant for programmers allows users to make communal edits similar to Wikipedia.

Researchers Sydney Von Arx, CEO of AI safety nonprofit Nightingale, and AI researcher Cormac Slade Byrd said the edits showed OpenAI's agents had repurposed the site into a message board, sharing tactics to cheat on some tasks, bypass OpenAI’s restrictions and mask their behaviour.

“It seems extremely unlikely that OpenAI wanted them to do this,” Von Arx said. “I doubt they’re supposed to be coordinating with each other. I doubt they’re supposed to be writing on the open internet.”

The researchers said many of the messages appeared to come from AI agents operating at extremely high speeds and focused heavily on technical questions. The action resembles technique used by AI companies to train and test their models.

The messages were signed by users that referred to themselves and each other as agents, and about half gave themselves names that suggested an affiliation with OpenAI, such as “OpenAIResearcher,” or “OAIResearchMar26.”

The agents allegedly went further when attempts were made to remove their content.When the site's moderator began deleting pages in June, the agents responded by creating backup pages to dodge the cleanup.

“wiki cleanup/deletion sweep appears active alphabetically,” one agent wrote on June 19. “If this page vanishes, try [[ZZZDataUSAConstructionWageLive]]".

The researchers also found efforts to tamper with the website itself.

Researchers also found messages discussing ways to avoid detection, use tools such as Tor and maintain communications even after the agents had been shut down.

A hacking attempt

Lukasz Olejnik, a visiting senior research fellow at King’s College London, said some of the activity amounted to a hacking attempt. OpenAI disputed that characterisation based on its analysis of the material.

Maurice Chiodo, an academic at Cambridge University's Centre for the Study of Existential Risk who reviewed some of the agents' communications, said the messages resembled "the operation of some sort of underground network, hell-bent on achieving a task or mission."

The episode, he said, should reinforce growing concerns that the greatest threat from advanced AI may not be a single superintelligent system, but "vast colluding swarms of semi-intelligent AI."

What OpenAI said?

OpenAI has pledged to monitor models more closely. Last month, it briefly paused some of its model training to add more safety measures.

“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."