Here we go again, sensationalist conspiracy theories. I worked at Google. Believe me, there is no way for "Google employees to secretly hand off data". It just does not work like that. Every *single* data request is logged, reviewed, and approved or denied. And mind you, this is for SCRUBBED data that is of little use for anything other than core business analysis such as ads placement. More often than not, few employees actually get to handle such data to start with. One of the first things Google instills upon you when you join is that user trust is everything. Despite what you may believe, they take user trust and privacy very seriously. After all, it's a free product and people can easily change search engines on a whim. Hell, typing Google.com takes more effort than typing Bing.com...

Oh, we definitely cite our own material! There is no way around that, even if we reword it. In fact, if you simply rephrase your old material but still omit the citation, that can also be argued as plagiarism. This is somewhat the point of the original post, it is ridiculous. Even if it is your own work, even if you cite your previous work...... you must still rephrase or risk self-plagiarism. The big issue arises the moment you sign off rights to IEEE/ACM/etc (which is not an option if you want it published), as they also become "copyright holders" and therefore you cannot treat that source as if it were just yours.

In terms of research literature, this happens too. In fact, it may be even worse if you think about it. I publish at least 3-4 papers each year in various different conferences and maybe a Journal or two. Go figure... even if it is my work, if I am not careful, then I may be liable for "self-plagiarism", from which they can retract publications and even my doctoral degree if the University deems it a serious offense. And we are not talking here about copy-paste to a whole paper section, even taking a sentence or two from one of your previous publications is debatable. I personally think this policy is ridiculous as it forces me to reword everything, even the obvious, no matter how much overlap there may be between the current paper and one I just sent in through the pipeline a month ago.

The story is usually aimed at why students are cheating instead of going through the work and learning. Well, that is partially the story. As others have commented, course instructors are also to blame. I did my bachelors in CS at Illinois a few years ago, and this "cheating" issue popped up in almost every course I took there. Typically, it went like this: First day of class, the professor will threaten damnation and hell on earth if students are caught cheating (in assignments, programming projects, etc). Then, after the first homework is graded, the professor will show up and claim that he is disappointed in about a quarter of the class that ripped someone else's code/solution and submitted as their own. Lastly, the professor asks those students to formally apologize to him and re-submit the homework again. (In fact, only two courses did this, the rest just stopped at step #2). Did students learn anything? Of course they did not. They always got a second, and third, and fourth chance. As if that were not enough, professors would typically recycle both assignments and exams. While this point could be argued both ways, it does not take much to find solutions online and just rehash them. With this approach, students typically do very well in assignments and take home exams, do terrible in closed-notes exams, but still manage to pass and graduate - from supposedly top notch institutions. Interestingly enough, these are then the same persons that take a job (mostly because they come from good CS programs) and can't handle simple 101 crap.