Most interpreted languages are "safer" in this regard as they don't leave the memory management up to their users. The interpreter itself could still have such

Most interpreted languages are "safer" in this regard as they don't leave the memory management up to their users. The interpreter itself could still have such bugs, but that's a comparably much smaller attack surface and has far more eyes on it that any random person's code, so these are less frequent. And it's not just fully interpreted languages - any language that manages memory for your and doesn't let you mess with it yourself (like Java, C#...) will be safer, assuming the compiler and runtime are secure.