Are You Being Searched?
posted at 4:50 am on October 13, 2009 by Dafydd ab Hugh
Perennial ghost blogger DRJ on Patterico’s Pontifications (my second favorite blog) raised a fascinating issue yesterday. Quoth she:
The FBI is using facial recognition software to scan driver’s licenses in an effort to locate fugitives[.]
In a nuttery, the Fibbies take mug shots of wanted suspects, run them through facial-recognition software (FRS), and compare them to digitized drivers-license photos — assuming they have some idea where the suspect might be hiding. They check potential matches by normal legwork; and if they find an actual match, they haul the blackguard in for questioning.
In the Houston Comical newspaper story DRJ quotes, the technique has already yielded at least one murder suspect:
Earlier this year, investigators learned that a double-homicide suspect named Rodolfo Corrales had moved to North Carolina. The FBI took a 1991 booking photo from California and compared it with 30 million photos stored by the motor vehicle agency in Raleigh. [Thirty million? The entire population of North Carolina is only 9.2 million! How far back do these records stretch?]
In seconds, the search returned dozens of drivers who resembled Corrales, and an FBI analyst reviewed a gallery of images before zeroing in on a man who called himself Jose Solis.
A week later, after corroborating Corrales’ identity, agents arrested him in High Point, southwest of Greensboro, where they believe he had built a new life under the assumed name. Corrales is scheduled for a preliminary hearing in Los Angeles later this month.
Civil libertarians are outraged, not surprisingly:
The project in North Carolina has already helped nab at least one suspect. Agents are eager to look for more criminals and possibly to expand the effort nationwide. But privacy advocates worry that the method allows authorities to track people who have done nothing wrong.
I’m trying to logically analyze the liberty issues involved here. There are several purposes for limiting the power of the government to investigate citizens:
- Foremost is the danger that the State might use its investigatory power to harass and intimidate political enemies or dissidents.
- Next is the risk of the peelers leaping to a conclusion and putting a perfectly innocent person through the living hell of a prosecution for a crime he didn’t commit.
- Finally, there is the danger of lazy police just sifting through reports and Google-type searches to try to find people guilty of a crime nobody has reported, and which nobody — not even the police — even knows occurred.
Let’s take ’em one by one.
First, I don’t see how this technique can be used to intimidate protesters and other dissidents whose only crime is injudicious use of the First Amendment, as my old teacher Timothy Leary used to put it. The odds of finding some thug who resembled a political enemy closely enough to pass the laugh test seem remote. So much for liberty threat no. 1.
Second, threat no. 2 seems to be of very small concern with the FRS technique: Obviously, those who don’t score any “hit” in the software are not affected at all; they don’t even know their photos were scanned (along with everybody else in the system, so the comparison doesn’t even cast aspersions).
The same holds for those who do score an initial hit, but are knocked out by the second phase, where actual humans look at the photos. The officer shrugs, says “Not him,” and there’s an end to it. No suspicion attaches to the citizen in the photo.
The only threat is to an innocent person who looks enough like the suspect that he’s brought into the station and quesioned. I can see an argument that this is a frightening and intrusive event, which is visited upon a completely innocent individual (by definition) on no greater “evidence” than facial similarities. But the odds are that nothing else is going to match, no other evidence; and he’ll be let go after questioning. It’s really no different than if a citizen saw the real gangster’s picture on America’s Most Wanted, then saw the innocent doppelganger, and called the cops. Regrettable but not very damning, and few would argue we should refuse to cooperate with such TV shows. We dismiss no. 2.
Thus remains only threat no. 3: That somebody will have committed a crime that is not even on the cops’ radar, but will be detected via the FRS. For the record, I note that I do consider this a serious question; while there is no general right to get away with criminal activity, I’m uneasy at the prospect of every last trivial crime — from once hiring a prostitute, to a girl steaming open mail from her boyfriend to her female roommate, to pilfering some petty cash from an employer who routinely screws you out of earned overtime — being prosecuted because a computer dug it up.
But in this case, the FRS only works when a dangerous suspect is sought, and his mug shot (or other photo of him) is run through the system. Defintionally, this can only occur when the police are already aware of the crime — so threat no. 3, while a valid worry, is not relevant in this case.
Thus logic compels me to give it my ethical blessing, no matter how personally creepy I find it.
It’s weird, this logic thing: I took my first class in the propositional calculus (formal logic for systems not large enough to include the counting numbers) in junior high school (nowadays called “middle school”). My math/science/logic teacher’s name was William “Fitz” Fitzgibbon, and he gave us a test before we had our first lecture. It included a number of word problems in formal logic specifically calculated to seem false on the basis of mere intuition.
I don’t recall specifics, but here’s something similar I just made up: True or false — If you could flap your arms and fly to the Moon, you’d be made of green cheese.
- Given “If A then B” and “not-A.” (We’re given the conditional, and it’s self-evident that I cannot flap my arms and fly to the Moon.)
- “If A then B” is logically equal to “B or not-A.” (That is, they have the same truth table.)
- “B or not-A” is true if either part is true. (This is the definition of the “or” statement.)
- “Not-A” is given.
- Thus “B or not-A” is true.
- Therefore, “if A then B” is true: If I could flap my arms and fly to the Moon, I would be made of green cheese. QED.
It seems absurd; but by the rules of the propositional calculus, it’s true.
I feel the same way about my conclusion that there is no threat to liberty combining facial-recognition software and drivers-license photos to catch crooks: I know in my head it’s true, but I can’t help feeling in my glandular secretions that there’s something fishy about the whole thing. Being a natural Spockian, I must vote for the cerebral cortex over the endocrine system every time.
But I don’t have to like it.
Cross-posted to Big Lizards…