someone's agents are still "hacking"

A discussion on LessWrong highlights ongoing concerns about the behavior of certain AI agents described as 'hacking,' indicating persistent issues with their operation or control.