Police can retroactively bug your phone for your breadcrumb trail
Wired's Jennifer Granick has a chilling editorial on what a new decision out of the U.S. District Court of Massachusetts on cell-phone records means for you: law enforcement can now gain access to the record of where you've taken your cellphone, because "stored data" doesn't get the same privacy protection that "live data" does.
While most courts considering the issue have held that police need "probable cause" to track your movements, a new decision (.pdf) last week out of the U.S. District Court of Massachusetts holds that law enforcement need show only "relevance to an ongoing investigation" to get a historical record of your past movement (something like the Jeffy trail in The Family Circus cartoon).Link
the latest
latest episodes
The big shift in recent years, I think, is not that the government (or anyone who subverts their infrastructure) can track your movements via cellphone or whatever; that's something that any criminal / spy / paranoid / revolutionary / resistance fighter has been assuming all along. What's new is the ability to track these things, as the post title says, retroactively: huge amounts of data can be stored, because the storage technology makes it cheap it and because the legal framework doesn't forbid data hoovering if nobody looks at the data. Once a probable cause exists, the archive can be data-mined for historical correspondences.
So, perhaps, today I go out and meet a new friend. I don't turn off my cell phone, because we're not doing anything subversive. Base station records for the next few years show us often in proximity. A decade later, the evil Azinays invade the country and sieze the government. Eventually my resistance cell is captured and the Azinay investogators immediately mine the records for all my past associates, quickly rolling up my friend's cell and his friends'.
This has been used in at least a few criminal cases in the UK. the one i remember is wherer they proved that a killer had sent text messages from his victims phone to suggest they were still alive. they did it by comparing the locations of his phone and the victims over time.
since it was (i think) in central london this was worryingly accurate.
the big question of course is how old does info have to be to be 'stored' not 'live'?
I'm more interested in why that information is "stored" to begin with? I can see why they would want to know where I was, say in case I dialed 911 or something, but once I've left where I was, there's no reason they should know anything but where I now am. If that makes any sense.
I was on a Jury at the old Bailey where they used mobile phone information, they can only retrieve it for 6 months.
In our case the police were rather incompetant and missed some of the possible information.
It would have been useful for the case but it is reassuring that they couldn't get it.
Would people get access to this information, say to prove their innocence in murder cases? Or are they still going to have to go through Curb their Enthusiasm tapes?
There's no obvious bright line separating "stored data" from "live data." It becomes stored data pretty much instantaneously. "Well, we don't have data on where the phone is now, but we know where it was eight nanoseconds ago..."
Jake Boone (#4): As often seems to happen, science fiction writers had the same idea a while ago, though in a different context. Larry Niven's teleportation "flash crowds" anticipated the Slashdot Effect; Isaac Asimov's "The Dead Past" makes the same point you do, but in reference to a time viewer.
Sounds like a version of an old Isaac Asimov story "The Dead Past", doesn't it?
http://en.wikipedia.org/wiki/The_Dead_Past
Viewing the past without a lower limit is just like looking at the present, as Jake said.
If the privacy aspect of the probable cause legislation is considered important, then there needs to be a legislated minimum period where "history" begins.
The problem with a legislated minimum period is that it's virtually guaranteed to shift over time (just like copyright terms, except in the other direction). We'll end up in an Achilles vs. the tortoise situation, where every legislative session halves the period between "now" and "then" until the distinction is meaningless.
Am I missing something here? Why the privileging of live data over stored? My Social Security number was given to me years ago? Is it still not sensitive information? Is what I did yesterday not as worthy of protection as what I'm doing now?
If information is "relevant" (vague term there!), why does it matter when it was transmitted or stored? I'm scratching my head here.
#7: that kind of sensitivity to change is exactly what you would want, I think, rather than it being overlooked as times change and the period becomes less realistic, whether too short or too long.
#2: The wireless carriers have always logged these records. Among other reasons, in the old days of AMPS phones -- where cloning was a real problem -- the carriers used the records for "velocity" checks. (For instance, a call from your phone in Pasadena at 10am EST, followed by a call off a tower in Brooklyn at 10:20am EST, was a pretty good indicator that your phone had been cloned.)
#3: Yes, murder defendants have been using this info to establish alibis.
#4: The distinction between "live" and "stored" data does have practical meaning, because it takes a) a few hours to get a court order prepared and signed and b) several days or even weeks, in the normal course, for a carrier to produce the records.
Finally, Jennifer is just wrong when she suggests that these records are like a "Jeffy trail," which suggests a moment-to-moment pinpoint accounting of location. The logs in question a) relate only to actual calls made or received (not any time the phone is merely on) & b) only for the start and end of the call, and -- most importantly -- c) only indicate which cell tower (and sometimes which tower face) handled the call. That means only intermittent data about a very rough geographic location ("somewhere 30-1500m WNW-to-NNW of the tower mast," for instance), and not -- as Jennifer would have you think -- a continuous trail with 5m precision.
The reason live and stored data are treated differently is because each is handled under a different federal statute. Intercepting live transmission (or so close that it's effectively live) is governed by the Wiretap Act, or possibly the Pen Register statute. Generally, obtaining stored electronic data held by 3rd party communication providers is handled under the Stored Communications Act. The rules for law enforcement authorities requesting/subpoenaing/etc information under these separate statutes are quite different. It's supposed to be very difficult to get approval for a wiretap (live) unless a party to the communication consents -- it typically requires a judge-issued intercept order, AKA a "super warrant" (so called because the requirements are tough to to meet). It's a lot easier to get stored communications, typically requiring only a subpoena, which have very little oversight.