That is a risk, and it may be exploited some day. It is also one that has been on my mind since I first heard of the predator test flights.
Every system has it’s potential risks, but they are weighed and mitigated against the impact and likelihood of those risks being realized.
The benefits are huge while the risk seems to have been successfully mitigated so far. I wondered why it was hard to break the security until I found out they get their instructions from satelites. Apparently they are selectively blind to anything coming from the ground so the key to hacking them is taking out the satellites. That may be off in the future, and by then we may have measures to deal with it.
Then again, you never know.
I could see where there would be mitigating factors when you have a limited number of drones in theatre. However, if they go to a more wide spread use including combat aircraft, perhaps those mitigating factors may not provide the same level of security? I’m skeptical and would prefer to have someone in the combat aircraft using all their senses on the battlefield - something that cannot be done remotely.