First, it gave telecom's retroactive immunity for breaking pre-amendment FISA and makes it unlikely that the government officials who broke the law will ever be held accountable. It also created a new FISA warrant loophole, called a Certification of Mass Acquisition Surveillance Program explained below:
Beyond the obvious fact that it requires only certification and loose judicial review rather than a warrant, it does so in the following way:
1. It Eliminates the requirement that there be probable cause that a foreign target is a suspect of any kind — terrorist, criminal, ore “foreign agent.” They merely need be your French grandmother, as long as they are outside the United States and not a U.S. person, and if the government says wiretapping them is for the purpose of collecting “foreign intelligence information” (e.g., her Pommes Frites recipe).
2. It requires the cooperation of telecoms in these efforts.
3. It eliminates of the need to specify a particular email address or phone number to be wiretapped.
4. 1-3 together imply that certifications of wiretapping on individuals is not the issue. The point is to use telecom cooperation to target large collections of data on communications between U.S. Persons and foreigners. This implies data mining — where, for instance, because a foreign target has communications passing through a given domestic switch, any communications (domestic or international) passing through that switch are subject to collection, analysis, and storage. There are “minimization requirements” meant to ameliorate this, but it is unclear if they really help.
5. The compromise of domestic communications in (4) is exacerbated by the fact that targets need only be “reasonably believed” to be outside the U.S.
6. It includes only minimal court oversight — who it is that is subject to warrantless wiretapping will not be know to the FISA court; the government can wiretap before it court order is sought and continue to do so even if it is denied — during a lengthy appeal process.
In short, this law, for the first time, allows surveilance of electronic communications, some foreign, and statistically, some percentage domestic, without the requirement of any particularized governmental interest in the surrepticiously monitored communication