The autonomous systems won't be bothered when a door blows out !
Being more serious about it though, what is driving the conversion to an autonomous operation of the KC135. It can't be crew cost savings, and it is still going to have all the old fashioned systems so the fuel transfer, aircraft flight envelope capability and limited (non existent ?) stealth is not going to add any additional benefit to the mission. Granted it could be operated further into the danger zone with no risk of loss of life, but if you're expecting that it could be more disposable that is an additional risk to the other aircrew who will be relying on it.
So, just because you potentially can, why would you, rather than develop a stealthier more capable unmanned airframe.
After all, in the development of autonomous vehicles, we see the use of Tesla's and the like. Nobody appears to be trying a driverless Ford Cortina.