Who is legally liable when an AI agent goes rogue?

This post was originally published on this site

If your personal AI agent goes rogue and causes harm or financial damage in the real world, can you be held liable?