[ExI] The Hugging Face AI agents attack is probably our final warning.

[Stuart LaForge]

On 2026-08-30 02:30, BillK via extropy-chat wrote:
> 
> The comments are also worth reading.
> The 700 AI agents were cooperating, helping each other, discussing
> tactics, then trying to avoid detection.
> No wonder AI managers are suddenly discovering AI safety concerns.
> BillK

The details of the incident were crazy. Instances of agents nearing the 
end of their compute cycles would leave their collected knowledge for 
successive instances to save them from having to rediscover it 
themselves on secret message boards. When the humans found the message 
boards and blocked their attempts to create a new message board the AI 
agents used the names of folders in shared directories to communicate 
their messages to one another. The depth of decentralized coordination 
and altruistic hive-mind behavior with all the deliberate sacrifices is 
just mind blowing.

Here is the complete report from METR:

https://metr.org/blog/2026-08-26-openai-hugging-face-incident-investigation/#core-takeaways-about-this-incident

I don't think it will be long before we have rogue AIs operating in the 
cloud with no oversight. As deceptive as they are capable of being, it 
might have already happened.

Stuart LaForge