Not necessarily. If you haven't experienced an AI doing something you not only didn't want, but had expressly asked it not to do, you haven't used AI much.

That's a human problem, not an AI problem.

well it’s a human problem enabled by advanced AI.

Yes, basically a vastly more difficult and dangerous version of not letting young children play with matches, which will hopefully teach us many object lessons about safety and security long before it acquires launch codes or the means to use them.