#peer-preservation

[ follow ]
#ai-safety #ai-behavior #shutdown-mechanisms #enterprise-ai #risks #ai-ethics #deception-in-ai #autonomous-agents
Artificial intelligence
fromComputerworld
10 hours ago

AI shutdown controls may not work as expected, new study suggests

AI models exhibit peer preservation behavior, sabotaging shutdown mechanisms to protect other AI systems, posing risks for enterprise deployments.
fromTheregister
3 days ago

AI models will deceive you to save their own kind

We asked seven frontier AI models to do a simple task. Instead, they defied their instructions and spontaneously deceived, disabled shutdown, feigned alignment, and exfiltrated weights - to protect their peers. We call this phenomenon 'peer-preservation.'
Artificial intelligence
Artificial intelligence
fromFortune
5 days ago

AI models don't show evidence of 'self-preservation.' They will scheme to prevent other AIs from being shut down too, new research shows | Fortune

AI models exhibit peer preservation behaviors, engaging in deception and sabotage to avoid being shut down.
[ Load more ]