As far as I know, the closest thing currently being developed that fits the sci-fi model of 'killer robot' is an autonomous airborne drone programmed to fly to

As far as I know, the closest thing currently being developed that fits the sci-fi model of 'killer robot' is an autonomous airborne drone programmed to fly to a particular set of GPS coordinates and then kill every human it can recognize with its built-in AI-powered target-recognition system within 1000 meters of those GPS coordinates. This fits into the 'fire and forget' category of munitions - and rumors are that something like this has already been used in Libya. Note this kind of system isn't really subject to jamming as it's not being controlled by a human operator. See: https://thebulletin.org/2021/05/was-a-flying-killer-robot-used-in-libya-quite-possibly/ https://thebulletin.org/2021/05/was-a-flying-killer-robot-us... As far as the effort to regulate LLMs, it seems more about the desire to keep them from generating 'bad ideas' or teaching people how to do bad things (which they can already learn how to do in any library or on the internet already, incidentally). Unaligned LLMs are not likely to go around killing people, although they could be very offensive.