Engineer Questions Liability for Agentic AI Systems
Engineer Andrew McCalip questions ownership of liability for self-managing agentic systems controlling resources.
TLDR
Andrew McCalip posted that he holds strong conviction self-managing agentic systems will soon control large amounts of resources. He asked who owns liability for their actions and listed three candidates: the host of the model, the trainer, or the person who deployed the system. McCalip works as an engineer at Varda Space Industries building spacecraft and previously co-founded Cosine Additive, which GE acquired. The statement appeared in a public post discussing future deployment of such systems.
Combined views
3.9K
1 Source, first seen 28d ago