These are target drones, and unlikely to be armed with anything or operated near high-value infrastructure susceptible to the relatively minor impact damage of which a QF-16 might be capable.
And why would you have a small very expensive fleet of fast moving drones that can carry bombs? Targets/training only? Wouldn't that be wasteful? Wouldn't those be expensive to just send in for recon or shooting down a plane or even a bombing run that they could use other manned aircraft for? What would they be best suited for?
"...the Pentagon is also modifying Turkey’s current fleet of F-16s to carry the bomb.":
http://www.f-16.net/forum/viewtopic.php?t=22037
I'm not against the USAF trying to improve its defense.
But, if a remote-controlled nuclear-armed jet were controlled by an AI, or by a terrorist, that would be a big problem.
I understand that it's expensive to find and train good pilots, and it takes a long time. They are a limited resource. So, I can see why you'd want to replace them with drones- you save a lot of money, could expand the fleet more quickly and with less expense, eventually.
But, that expense is a safeguard. It keeps really bad things from being able to happen. We shouldn't be making anything nuclear-capable more easily remotely-controllable. And if the US does it, others will. Even if the US were to be completely safe about it, other states might not be.
As throwanem said these are target drone which we have been doing with the F-4 and other planes for a long time. They have gone beyond their service hours for the airframes.
Nobody loads nukes on a target drone. Nobody uses target drones near civilians.
http://www.boeing.com/features/2013/09/bds-qf16-09-23-13.pag...
They do carry a bomb, though- it's not meant for combat, but could be abused for such to make the entire plane a bomb carrying device, if it were taken over:
"Granted it is removable, but they also carry, as it was described to me, a 2000 lb class (or maybe more reasonably it was 500#, but the 2k figure seems to stick out in my memory for some reason) HE warhead in the spine for range control/destruction purposes if anything goes awry. Long story short, it would take a lot more money and a lot more testing to ever make them combat capable."
Indeed, a firsthand source [1] describes the QF-16's range safety package thus:
> A small ball of flames burst from the middle of the aircraft, followed by thick black smoke, but no sound. The sound followed soon after with a deep reverberating boom.
> The extent of the damage went undetected at first, due to the amount of smoke billowing from the wreck. Once it cleared, it revealed the F-16 had been split in half between the cockpit and the wings.
Something on the scale of a 500- or 2000-pound bomb, as this random forum commenter suggests, would've scattered pieces of aircraft all over the hardstand, in which it would also have left a hole.
[1] http://www.defense-aerospace.com/cgi-bin/client/modele.pl?pr...
Why do people get so hung up on the idea that a superhuman AI can't gain access to weapons if there are no weapons equipped for direct remote control? Why expect a superhuman AI to have difficulty suborning humans through deceit? I mean, don't get me wrong - I tend to think that whole avenue of existential threat analysis owes much more to science fiction and the seductive nature of apocalypticism than to anything in reality. But if those are your priors, this seems a very odd oversight.