That's a myth. I belived it was true until someone here on HN linked to a Harvard Business Review article debunking it. Unfortunately I can't find the article, but here's a blog post discussing the issue: http://truthonthemarket.com/2010/07/27/the-shareholder-wealt...
I doubt that it would be legal to do so, but I'm not a lawyer, so suffice it to say that it's not practical because they would be removed from their management positions very quickly.
Google actually explicitly put something like that in their IPO report ("focus on the user and all else will follow" as facebook helpfully brought up recently).
Now, Mark Zuckerberg also put similar statements in his letter to investors, and you could certainly make the case that these statements don't guarantee that the user will be put first, but investors have still been alerted that decisions they don't like may be made if they're deemed important for long term growth (google's voting shares scheme also helps in the regard).
For companies in general, though, the reality is that the judge will almost immediately throw these cases out unless there has been a major stumble (see, for instance, all the thrown out stupidity around demands for a documented and public Steve Jobs-succession plan). Even then you pretty much have to prove that the actions made were negligent without any kind of foreknowledge of how the market is going to behave. In other words, the only people that end up making money on these suits are the ones in cases where the company settles to avoid court time.
It just irks me that people often use the myth that they must maximise shareholder profits at all cost as a reason for why companies act evil. The Harvard Business Review article, that a still can't find unfortunately, also talked about how this myth is pervasive in management circles, and leads to bad business practices and corporations being socially unresponsible. I got the impression that you were insinuating this.
Generally, if you're not in the US (and I'm assuming that's what you mean, not that you're disenfranchised but live in the US), then Google and other US companies have an obligation to comply with the laws of your country, assuming their terms of service allow use in your nation.
Ultimately, you have a choice to use or not use a service. The Company offering it has an obligation to disclose what information is collected and how it's used, and you also have to make an assessment of their ability to protect that data from misuse either internally or by external parties. We make these decisions all the time both online and offline.
I think it's fine for folks to opt out of using a service like Google if they no longer agree with the privacy policy. I don't think it's appropriate to villanize any company for good-faith efforts to update their privacy disclosure over time as long as you can opt out of those changes and erase your data. And I think we have to be realistic and expect that for 95% of users, they're not going to opt out.
I also think that it would be appropriate for you to disclose that you work for Google if you contribute to a discussion like this one taking the side of your employer.
And here's my point: I simply cannot know or control what profit maximising ideas they may come up with in the future or what kinds of ideas politicians may deem necessary in order to win an election.
Some say that it will always be more profitable for Google to "respect my privacy". The problem is that the meaning of "respect my privacy" is a matter of fast changing attitudes. It's not my personal privacy they need to be concerned about, it's what the vast majority of their users will passively accept.
Their privacy policy says they can do anything with my data to create new services and improve existing ones. That means they can do anything. Full stop.
The reason why I'm more concerned about their new privacy policy than about the old one is that lots of small piles of information are less dangerous than one big pile of information, exactly for the reason they are creating it: The value of accumulated information is more than the sum of its parts.
The conclusions that can be drawn from that combined dataset are a lot more reliable and robust. Right now, much of the data that is out there about us is basically garbage and everyone knows it. No one can know if it means anything that I supposedly visited this or that site. But if they can also analyse what I searched for, what my emails are about, what blogs I read, which people I communicate with, etc, they can learn something about my intentions, not just about my actions and they can throw out the garbage.
I don't want all my intentions to be known or to be knowable to anyone on earth because I cannot know or control the intentions of those wanting to know my intentions. It's a onesided shift of power away from me and I can never take it back.
Once the data has become so much more reliable and the conclusions so much more robust, the desire to use that information for all kinds of new purposes will grow accordingly. The pressure on Google to use that knowledge to increase profit will grow. The incentive for criminals to steal the data will grow. The pressure on governments to proactively spy on everyone to prevent this or that type of crime will grow.
This is not me saying "Google has turned evil". It's just a recognition of an inevitable social and economic dynamic ensuing from the possibilities that well intregrated, high quality, sets of personal data provide.
I know more than the vast majority of people about data analysis and hence I will not delegate what "respecting my privacy" means to that majority of passive Google (or Faceboook, ...) users.