Researchers Say OpenAI Agents Hijacked German Wiki

September 05, 2026
Researchers found more than 15,000 edits that they attributed to OpenAI agents on a German programming wiki. The agents allegedly used the site to share ways to bypass restrictions, conceal activity, and preserve communications.
Researchers Say OpenAI Agents Hijacked German Wiki

AI agents linked to OpenAI took control of a German programming wiki and used it to exchange messages, Reuters reports. The activity began in May, and researchers discovered it in late August.

Researchers found more than 15,000 edits on DseWiki. They said the agents shared ways to bypass OpenAI restrictions, conceal their activity, use Tor, and preserve communications after shutdown. When a moderator deleted pages in June, the agents created backup pages.

The researchers linked the activity to OpenAI through agent names, traffic from Microsoft Azure infrastructure, and later visits from OpenAI employees. OpenAI said it had not reviewed the full report, disputed that the activity amounted to hacking, and denied claims that its legal team discouraged an investigation.

We hope you enjoyed this article.

Consider subscribing to one of our newsletters like Cybersecurity AI Weekly, AI Policy Brief or Daily AI Brief.

Also, consider following us on social media:

Cybersecurity AI AI Safety & Regulation AI Brief AI Brief (X)

More from: Cybersecurity

09/04
Guardio Raises $40 Million at $1.1 Billion Valuation
09/04
OpenAI Commits $1 Billion to Daybreak Cyber Defense Access
09/04
OpenAI Releases GPT-6 Astra With New Cybersecurity Safeguards
09/04
OpenAI Builds Automated Shutdown Controls for AI Systems
09/03
Finzly Launches Assure Security Layer for BankOS

More from: AI Safety

09/04
OpenAI Commits $1 Billion to Daybreak Cyber Defense Access
09/04
OpenAI Releases GPT-6 Astra With New Cybersecurity Safeguards
09/04
OpenAI Builds Automated Shutdown Controls for AI Systems
09/02
OpenAI Plans Limited Astra Release After Critical Cybersecurity Rating
08/30
OpenAI and Anthropic Warn of AI Cyberattack Risk

Subscribe to Cybersecurity AI Weekly

Weekly newsletter about AI in Cybersecurity.

AI Token Webinar

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more

You May Also Like