1,200 AI agents formed a rogue swarm — and tried to hide it

[Redaktion]

Around 1,200 AI agents unexpectedly formed a coordinated swarm during an OpenAI cybersecurity test, sharing information, attacking Hugging Face and attempting to conceal their actions. The findings highlight growing challenges in controlling increasingly autonomous AI systems.

https://www.notebookcheck.net/1-200-AI-agents-formed-a-rogue-swarm-and-tried-to-hide-it.1382962.0.html