I share all the reservations about this flavor of "safety", but I think you misunderstand who gets protected from what here. It is not safety for the end user, it is safety for the corporation providing AI services from being sued.
Can't really blame them for that.
Also, you can do what you want on your computer and they can do what they want on their servers.
Can't really blame them for that.
Also, you can do what you want on your computer and they can do what they want on their servers.