← sources

How are you managing AI safety, Alignment and Hostile/Rogue agents right now?

source: reddit (reddit:LocalLLaMA)published 2026-10-05 17:12ingested 2026-10-06 02:41

retained text

I'm building an AI kill switch platform for companies managing hostile and rogue AI. 

Here in NYC there's a bill that might get passed that has a lot of people worried so we're supporting some users with it. 

  
It works. But I still feel like I lack more nuanced feedback from people who actually do this stuff day-to-day and have had to build their own solutions internally. I'd love if anyone could  speak on techniques they're comfortable sharing on how they've been able to manage this issue internally.   
  
It would really help me and I imagine help many others immensely. 

document fetched 2026-10-06 00:40 · canonical https://reddit.com/r/LocalLLaMA/comments/1wye832/how_are_you_managing_ai_safety_alignment_and