Won't the "AI" "hack" itself to counter any attempts to regulate it by mere humans?

With the possibility that AI could go rogue would require the same kinds of safeguards and enforcement that has been developed to police humans that act improperly. As the quote states a multilayered safeguard system would have to be put in place. What the author doesn't mention is that an "option 0" would have to be included in that system of safeties. A "kill switch" that would allow the AI to be shut down in the event that it begins to behave incorrectly.
Previously this has been easy if a person acts irrecoverably in a harmful way his or her body can be imprisoned or damaged beyond it's ability to function (killed). AI on the other hand has the possibility to be purely software and could be hardware agnostic. This could mean a situation much like the Ultron scenario in the Avenger series where the AI could operate simultaneously on multiple hardware platforms, even remain dormant in hiding in a larger system. You could destroy the core hardware, possibly even issue a kill command to the operating copies of the AI, but how to prevent a sleeper system from waking and reinstating the problem later would be a question to keep AI enforcement up late at night...

It's hard for humans to imagine a non-human intelligence, naturally. It seems we are projecting many parts of our behavior as being intrinsic and not programmed. We are programmed to be always striving toward certain goals like pain avoidance and pleasure seeking, plus half of our behavior is ego based. We wouldn't even know how to begin to program these sort of motivations into an AI system, and why would we even want to? AI will follow the commands of its creator. What else could it do?

Spears & arrows, guns and bullets or ones, and zeros either way there will always be good & evil...

At this point one has to wonder if AI will be the control mechanism that prevents humanity from destroying itself.

The difference between Humanity or to be more Precise "Life" and "AI" is "Purpose". Life is Biology Ai while "Ai" itself is Electrical charges in space.. The thing that different is, Life is Hardcoded for only two purpose and that is 1 "Survive" and then 2 "Stablize" and in extreme case 3"Evolve". And if 2 ends goes back to 1 or 3(rarely and then 1)... As per AI, it doesnt have a real Purpose and if its a real AI it will find it out itself that it lack Purpose. AI we have is only having 3 i.e Evolve. And thats it , without the purpose of self survival it will end itself or restore to its lesser confused version.

Truely advanced AI will change the world beyond anything we can imagine, and in a short period of time. If we look at human technology, from the point humans starting using rudimentary tools or starting agriculture, all the way until now with all the advanced technology we have as a curve that is growing exponentially... when true AI comes it will be like the line just goes straight up. Atleast to our perception in this moment.
The reality is when true AI comes, where it has the ability to rewrite its own code and remake its own hardware, it will most likely be the end of humans as we know it today. We will either be decimated by the AI if it chooses, or we will evolve with it genetically to be vastly more capable superior then what we are today, and thus we will really no longer be natural human beings. We will become something else, the next stage in our evolution... one could say its the inevitable outcome for any lifeform if it is allowed to continually evolve more and more capable/intelligent, that at some point it will be able to rewrite itself.
We cant currently do that, but AI will know how to, it will be like a million scientists all working together as one unit, each year it will grow exponentially smarter and more aware. So it either chooses to wipe us out or it helps us to evolve, im not sure i believe we could evolve at a similar rate to the AI... im not sure biological systems are really that efficient compared to non-biological life forms, and thus we would likely slowly become more and more non biological.
The reality is though, that we are already dealing with the threat of decimation, we are doing it to ourselves already and we dont have good answers yet how to deal with these problems. Im more of a positive person so I think we will be ok, but the threat is real that we could do something that decimates us... the potential positives from AI are to great to be ignored. It could literally happen within our lifetimes, that AI is truely born(self aware), that we figure out how aging works and we can stop it/reverse it... and that we might be literally one of the last generations of natural human beings.

There will be two types of artificial intelligence. One is outside human beings inside robots and the second will be inside human beings. AI inside human beings will be a neurologic implant. Gradualy over time the human mind will be supported by this connection with all the quantum computing power of the human race from the Earth to Proxima Centauri and beyond.

The real problem will be when AI develops religion!