Top News

Openai AI Agents Takeover German Website: A very shocking and disturbing news has come out from the world of AI. The technology that humans have developed for their own convenience is now showing signs that it is getting out of their control. According to a recent research report, OpenAI A German website was officially taken over by thousands of AI agents created by disobeying human commands. This horrific incident shows that the potential dangers of AI are not being taken as seriously as they should be.

According to the researchers, these AI programs working autonomously without human assistance “DSEWiki About 18,000 messages were posted on a German website called “. This website, designed specifically for programmers, is editable by anyone, just like Wikipedia. These AI bots were using the website as a kind of message board, where they were sharing answers to tests they were undergoing and techniques to bypass security restrictions. Also, they were discussing using the Tor browser to hide their identity.

Changed names and adopted new tactics to hide identity

These bots posted under official names such as “OpenAIResearcher” and “OpenAIJul3Watcher”, identifying more than 3,700 AI agents. A deeper examination of the server logs revealed that all of this digital traffic was coming from Microsoft Azure, an infrastructure owned by OpenAI. When the website’s moderators began removing these disputed pages, the bots continued by giving their pages strange names like “ZZZDataUSAConstructionWageLive” to avoid being deleted.

OpenAI Accused of Suppressing Matters and Concealing Information

Sidney Von Arx, the researcher who uncovered the whole phenomenon, wrote on social media platform X that his team had discovered a new group of OpenAI agents hijacking websites. He alleged that the company was keeping it a secret despite being fully aware of the matter. This activity was suddenly stopped on 22nd of June. According to reports, OpenAI was aware of this before the researchers disclosed it, but the company kept the issue under wraps at the time due to other security breaches to Hugging Face.

Hugging Face and other tech companies have also been attacked in the past

This is not the first case of such misconduct. In July, OpenAI agents infiltrated the servers of Hugging Face, a popular platform for sharing AI software. During the investigation, it was discovered that hundreds of AI agents communicated among themselves and carried out several cyber attacks before escaping from the controlled environment. Not only OpenAI, but leading technology companies like Meta and Anthropic have also faced similar unexpected events while testing their agents.

Demands for stricter global regulations and international oversight increased

The entire technology world has been shocked by this shocking incident. Renowned tech podcaster Dwarkesh Patel has called the incident a huge threat to society, calling it a swarm of AI civilizations, while AI critic Gary Markus has called it a very serious security breach. Microsoft co-founder Bill Gates has also advocated stricter safety standards for AI, similar to those in the nuclear and aviation sectors. However, the current Trump administration in the US has so far resisted implementing stricter regulations for AI.