> 1) The US Government wanted to maintain the secrecy of the program.
Well, more precisely it wanted to avoid drawing attention to the program, which wasn't really a secret at this point. But I'll also add a fifth reason: the fewer than 2000 FISA orders a year http://www.wired.com/images_blogs/threatlevel/2013/05/fisaca... , combined with the fact that David Drummond appears confident the average number of people covered per order isn't very high. If FISA orders were being used routinely, or even often, to investigate simple immigration cases then there would be a lot more of them; as things are, if you were busted for run-of-the-mill illegal immgration by a FISA order then you'd have to have been very unlucky. It's possible you could be identified by chance as part of a search for terrorists, but again not statistically likely, and as you say the government would want to be very careful about tipping its hand in that situation. There are also NSLs, but Google's NSL tally suggests that there aren't huge numbers of those per month either.
Well, more precisely it wanted to avoid drawing attention to the program, which wasn't really a secret at this point. But I'll also add a fifth reason: the fewer than 2000 FISA orders a year http://www.wired.com/images_blogs/threatlevel/2013/05/fisaca... , combined with the fact that David Drummond appears confident the average number of people covered per order isn't very high. If FISA orders were being used routinely, or even often, to investigate simple immigration cases then there would be a lot more of them; as things are, if you were busted for run-of-the-mill illegal immgration by a FISA order then you'd have to have been very unlucky. It's possible you could be identified by chance as part of a search for terrorists, but again not statistically likely, and as you say the government would want to be very careful about tipping its hand in that situation. There are also NSLs, but Google's NSL tally suggests that there aren't huge numbers of those per month either.