The NSA story of the day is found at Slate this morning. The key news in the piece is this sentence:
A former telecom executive told us that efforts to obtain call details go back to early 2001, predating the 9/11 attacks and the president’s now celebrated secret executive order.
The piece also goes into a lot of detail, some of it speculative, on the nature of what NSA has been doing in terms of data mining. The authors compare the NSA’s activities to Total Information Awareness, which intended to have strong privacy protections in its architecture:
Adm. John Poindexter, TIA’s creator, believed in the potential intelligence benefits of data-mining broadband communications, but he was also well aware of the potential for excess. “We need a much more systematic approach” to data-mining and privacy protection, Poindexter said at a 2002 conference in Anaheim, Calif., sponsored by the Defense Advanced Research Projects Agency.
Poindexter envisioned a “privacy appliance,” a device that would strip any identifiers from the informationâ€”such as names or addressesâ€”so that government miners could see only patterns. Then if there was reason to believe that the information belonged to a group that was planning an attack, the government could seek a warrant and disable the privacy control for that specific data.
Congress, with an assist from the privacy lobby, killed TIA in 2003. My reaction at the time was that this was a mistake, and that DARPA should at least have been allowed to test TIA and see whether data mining worked before killing it. In the hindsight of recent revelations, TIA now looks like the intelligence community’s effort to move away from the legal and ethical grey zone of NSA data-mining to an above-board “new normalcy” that could enhance security without diminishing privacy and civil liberties. And when TIA was killed, those efforts stalled.