That's sort of irrelevant, the creepy part is that the behavioural prediction models exist at all, and are controlled by an entity whose only responsibility or duty is to generate profit
Frankly, if they did that, they wouldn't make very much money. They can demand money from advertisers because they have an information advantage over everyone else. That information advantage gives them an incentive to protect your personal information. After all, if they leaked the information a competitor could use it to formulate competitive advertising profiles, which would drive down earnings.
Edit: I wanted to add a few more thoughts since I let it spin around in my mind a bit. I think if you followed the line of argument regarding Google's profit motive as a corporation, you could reach the conclusion that their users' privacy(that is, the distribution of their private information) is a prerogative of theirs to protect. Their viability as a corporation depends on their profiles being the most accurate, and on being under their sole ownership. As such, they are driven to protect them.
I would conjecture that the profit motive for their need to protect their information has led to their relatively robust consumer protection ratings by groups like the EFF[0].
--------------
Advertisers have been thinking about demographics since there's been advertising. Advertisers' needs have dictated the TV shows that do and don't see the light of day for half a century. Behavior prediction models have been around for decades, now they're just better.
Or that this progress should only be limited to state owned institutions?
Googles business model isn't based on doing something malicious like selling your private data to mexican drug lords, neither does Apple, Microsoft or Facebook.
The real threat is some kind of data loss (where Google has a pretty good track record) and secret wiretapping warrants issued by secret courts based on secret laws. (but even there it's not in Googles interest and they are forced by the state to comply)