from the the-redactor's-dilemma dept
Julian Sanchez has a fantastic post noting what a difficult job it must be to be in charge of figuring out what to "redact" in government documents that are being released to the public -- because
redacting certain information may actually reveal a lot more than if that same info had been left readable. As an example of this, Sanchez goes through some of the redactions found in the
recently released documents from various service providers in how they comply with law enforcement requests for info. He walks through some of the documents, and actually has his attention called to one redacted passage in a template for requesting info. What caught his eye is that the redacted section was the
statutory definition of basic subscriber info that could be requested by law. In other words, there's nothing secretive at all -- and, in fact, he notes that other similar documents include identical information that has not been redacted at all.
If the information had not been redacted, he would have skimmed over it without thinking. But the fact that a government official thought that the statutory definition needed to be redacted, actually
called a lot more attention to questioning why that passage had been blacked out. From there, Sanchez does some educated and reasonable speculation, to suggest that government officials may be collecting cell tower info, rather than GPS info, requiring a lower standard to request -- because they can easily get enough info from that to determine where a person is, even without the specificity of the GPS info. It's no secret, of course, that you can triangulate location via tower info, if you have info on multiple towers -- but, Sanchez points out that there are likely ways to get pretty close even with single tower info, and that can be requested at a much lower standard by pretending that you're
not trying to pinpoint exact location, and not even asking for the triangulated location info.
It is all speculation, of course, but it's fascinating that what sent him down this path in the first place was the simple decision by a redactor to redact basic information that is obviously already public. Sanchez surmised that the really salient piece of info is the fact that information can be requested while a call is in progress, rather than after it's done, which is what is used to determine a more precise location, if the subject is moving (and switching towers):
Now, did this possibility first cross my mind when I looked at these documents? No, not really--but thinking about this stuff breeds paranoia, and so a lot of possibilities cross my mind. The pattern of redactions above make me a good deal more confident that this is probably a popular method of getting moderately detailed location info on the "cheap" in terms of legal process. In the criminal context, anyway--for intel, who knows. They make it explicit in some of these documents that the Justice Department's legal position is that they can get realtime full-GPS with a mere "relevance" court order, but they go ahead and apply for that kind of tracking under stricter rules because they don't want to risk suppression. Probably they're less worried about that when they're operating under FISA pen/trap orders. But if this is right, they may be pulling a bit of a fast one on judges here. Because a lot of these applications to judges--and certainly the Justice Department's legal briefs in the cases where courts have been reluctant to approve tracking on such a loose standard--imply that this cell site/sector data, why, it's so rough and approximate that it vaguely counts as tracking at all. Certainly, at any rate, it's not so precise as to invade any sort of privacy interest. Except that for a target in steady motion, it begins to seem as though they can probably get a substantially more precise fix.
Filed Under: government spying, redacting, streisand effect