Your comment assumes that the weapon was one-point safe, which was not true for all US devices. I cannot remember the source but apparently for many years (thr

Your comment assumes that the weapon was one-point safe, which was not true for all US devices. I cannot remember the source but apparently for many years (through at least the 80's, I believe) many US weapons could have produced a yield if subjected to a fire. Suppose the trigger device providing the precise timing could misbehave at high temperatures. Or, in the same source I mentioned earlier, it was disclosed that there were wiring paths which could have melted together in a fire and cause the weapon to fire.