FISA comprises Subchapter I, Chapter 36, of Title 50 of the U.S. code - or alternatively, as 50 U.S.C. §1801 et seq. It was passed in 1978 to protect against the president using national security to surveil American citizens.
The big problem that is faced today with FISA is that the NSA is conducting communications surveillance that would ostensibly violate the Act. There are a lot of factors on how we got to this place.
Starting with the Act itself, §1801(f) defines “Electronic surveillance” to mean:
(1) the acquisition by an electronic, mechanical, or other surveillance device of the contents of any wire or radio communication sent by or intended to be received by a particular, known United States person who is in the United States, if the contents are acquired by intentionally targeting that United States person, under circumstances in which a person has a reasonable expectation of privacy and a warrant would be required for law enforcement purposes;
(2) the acquisition by an electronic, mechanical, or other surveillance device of the contents of any wire communication to or from a person in the United States, without the consent of any party thereto, if such acquisition occurs in the United States, but does not include the acquisition of those communications of computer trespassers that would be permissible under section 2511 (2)(i) of title 18;
(3) the intentional acquisition by an electronic, mechanical, or other surveillance device of the contents of any radio communication, under circumstances in which a person has a reasonable expectation of privacy and a warrant would be required for law enforcement purposes, and if both the sender and all intended recipients are located within the United States; or
(4) the installation or use of an electronic, mechanical, or other surveillance device in the United States for monitoring to acquire information, other than from a wire or radio communication, under circumstances in which a person has a reasonable expectation of privacy and a warrant would be required for law enforcement purposes.
Note that the NSA program would ostensibly fall under §1801(f)(2) because interception is being done in the U.S. Note also the differences between §1801(f)(1) and (f)(2). In (f)(1), in order to qualify, the surveillance must be of a targeted U.S. Person (defined as a U.S. citizen or legal alien) in the U.S. §1801(f)(2) on the other hand only requires that one of the people whose communications are being intercepted, is in the U.S. In other words, he may be an al Qaeda operative who entered this country illegally. Targeting doesn't matter. Being here legally doesn't matter. And the difference? §1801(f)(1) applies to interception wherever it occurs, whereas (f)(2) only applies to interception done w/i the United States.
Why is this relevant? Because technology has changed dramatically in the almost 30 years since FISA was passed. Back then, electronic international communications were either via underseas copper cable or by satellite. The underseas cables could be tapped offshore through induction, and communications satellites of the time were all in geosynchronous orbits - which are, by necessity, over the Equator, and, thus, not over the U.S.
Contrast this with the present situation. Most overseas communications today are via fiber optic cables under the ocean. But it is almost, if not totally, impossible to tap such other than at the switches. Why? First, fiber optics relies on the fact that the shielding reflects light. In order to tap this, it would be necessary to scrape some of this away, resulting in a significant signal loss. This is esp. critical under the ocean, with those long cable runs. Secondly, fiber optics utilizes a technique called Time Division Multiplexing (TDM). This means that thousands of calls are being time-sliced on a single strand at any one time. Worse, the call setup information (including who is calling whom) is carried "out of band", meaning that this information travels on separate channels (also time-sliced). Without the setup information, tapping calls is pointless. Worse again, the individual fibers are bundled into cables, and it is possible that the call setup information may even be traveling on another strand.
So, the logical place to tap international communications is at the switches, which, by necessity, use computers to tie the call identification information to the specific calls. This is the only place that makes sense, and the only place that is practical. The problem is that the only place in the world where we can assure ourselves of intercepting the bulk of communications is within the United States. Very roughly, around 1/4 or so of the traffic coming into this country routes through ECHELON participants (UK, Australia, and NZ). The rest would be lost if we couldn't intercept communications w/i the U.S.
So, technology has required that interception of electronic communications has had to move to w/i the U.S. The result is that it has moved from §1801(f)(1) to (f)(2), which, in turn, has resulted in the standard for something being defined as "electronic surveillance" being changed from only applying to communications when the person in the United States is legally here and is intentionally targeted, to communications with anyone in the U.S., regardless of who was targeted or whether they are here legally. In other words, if someone enters the U.S. illegally, then calls a targeted individual, such as Osama ben Laden, overseas, the conversation would fall under §1801(f)(2), and, thus, presumably require a FISA warrant. Ditto, of course, if OBL calls anyone in the U.S.