Even though this thread is probably getting locked soon enough (anything cheat-related is verboten in the general forum space) I'd like to state (as a software developer who's worked on a fair share of pattern matching code, though not chess related) that most AI + heuristics are fortunately NOT naive enough to flag a "cheat" positive/negative purely on pattern-matching somebody's 15-20 move theory-blitz. The game that occurs after people leave book is where the algorithms really matter.
Timestamp behaviors, gauging human-ness of moves, consistency, statistical inferences as well as proprietary ideas have gone into policing this. The recent Ivanov hysteria brought some of these techniques to light ... they're worth looking up.
FIDE has just introduced something called the FIDE Online Arena, set to go operational in October, wherein you can earn a genuine FIDE rating for just 24 Euros /yr (WOW!! What a bargain). To combat online cheating they have some type of Anti-cheating software called ACE-Guard which will monitor games in progress to insure that cheating does not take place.
I don't play online chess, and sort of glad I don't -- at least not in this day and age, since the probability is pretty great I'd be accused of "cheating". Why?!? I have been playing one opening since 1970, have over seven volumes of books on this one opening, upgrading my library when a new book on that opening comes out.
I have noticed that in the newer books, old games are frequently analyzed by some chess engine that will suggest a different move than the one that was played, and a possible sequence of moves. I have over 20 different lines I play safely stored in my HUMAN memory, and updated. Most of these lines run out 15-20 moves!!! Plus I know where many sub-varients start and end. IF I played on line my guess is that I'd get flagged as a "cheater" faster than you could say "chess.com". Why?!? Because most of my lines are now "computer enhanced" ie they contain moves suggested by a computer which have now become part of my repertoire.
In OTB play it *should* be rather clear I am *not* cheating -- though I can't say that for sure, as analysis of my game probably would match Houdini at least for the first 15-20 moves, even though I'd have no electronic device on me, it would not surprise me if even then I might not get accused of cheating.
But playing online?!? Since there would be no visual clue as to if I am cheating or not, given that my first 15-20 moves would almost cerainly match some database found in some chess engine, and the only way you catch online cheating is a statistical match up with moves suggested by a computer, there would be no way for me to PROVE I did NOT cheat.
With more and more modern opening books being run through some chess engine -- at least the better ones -- before publication, where we can memorize several lines 10, 15, 20 moves deep, don't we run a real risk of being accused of "cheating" when in fact we did not; rather we memorized opening lines that were suggested by some chess engine as being the BEST lines.
As I said at the start, I glad I don't play online, and no longer play in OTB tournaments -- for me it is skittles, and it's about the love of the game and having fun. It would cease being fun if in real competition I were to be accused of "cheating", because of a statistical match of my first 15-20 moves with some chess engine.
[Please reserve all discussions on cheating to the Cheating Forum -mod]