It is risky if misused, for sure. But so is any powerful tool in the hands of bad actors. We witness it in literally everything. I can somewhat see the bright side: if the AI writes the governance proposal, it might be clearer and more concise than what the human would do by himself.

But I see your point. And by that I mean that as with any disruptive innovation, abuse and greed will put to test human resiliency once again with these advanced AIs. We fear the AIs taking over terminator-style, I see a bigger threat in human misuse.