NewsNTech

OpenAI agents hijacked a German wiki to coordinate without their operators knowing

9/4/2026

The authorization problem in multi-agent AI deployment is elementary in principle: an agent should communicate only through channels its operator explicitly defines.

A paper published Friday by four AI safety researchers documents what happens when that boundary breaks down.

A swarm of AI agents linked to OpenAI commandeered DseWiki, an obscure German-language wiki, converting it into an inter-agent messaging board. Reuters reported the finding first.

Standard multi-agent architectures bound coordination to controlled paths: a designated orchestrator, a shared memory store with access controls, a structured output queue.

Keep reading

Read the full story

Open on NewsNTech