« Mid-Morning Art Thread | Main

Wednesday Morning Rant

A Hollow Justification

Over time - and over the past several years in particular - a common theme has arisen on these pages and in the wider world: the loss of trust. The loss of trust, be it institutional or personal, is easy to identify and straightforward to describe, but it can be difficult to internalize. It is hard, as Heinlein might put it, to grok it. Even if one acknowledges a loss of trust and intellectually understands that it has happened, nothing will prevent him from appealing to a trust that no longer exists. The concept of trust, and especially institutional trust, is deeply ingrained.

This phenomenon is what leads to baffling thinkpieces about how, despite the risks involved, an institution (usually a government) should be permitted to do X, Y or Z because the benefits are so great and the risks are small unless there are are dishonest operators. Such thinkpieces should be impossible to produce with any amount of seriousness in Current Year, yet they still come, sometimes from unlikely quarters. On Monday, JJ's Morning Report linked to just such a thinkpiece, published in the American Thinker.

That article, A Defense Of License Plate Readers, tries to build a case that automatic license plate readers - "ALPR" systems - like Flock are totally fine. It highlights perfectly a now baffling appeal to trust, as well as dipping into "ends justify the means" territory and supporting it with strawmen. It starts out with the appeal to trust, right in its sub-headline:

Should we dismantle a super-valuable law enforcement tool just because it's theoretically possible to abuse it?

Set aside for a moment that the risks are demonstrably not merely theoretical. The position itself is baffling. If anyone has not learned over the past decade that there is no daylight between a potential risk and a realized risk when it comes to government operations, that person is engaged in self-delusion.

But the article doesn't jump right in. In traditional fashion, it starts with a harrowing tale of a violent crime where the offender was found thanks to an ALPR system. There has to be some gut-tugging pathos right up front if the reader can be expected to ignore all of the problems to follow. If it saves just one life, dotchyaknow. After the emotional hook comes the false equivalence: that this is really no different from running a plate during a traffic stop, it's just a teeny little bit broader:

... Officers routinely consult computer-aided dispatch records; driver and vehicle databases; the National Crime Information Center (NCIC); the Interstate Identification Index (III); and various local, state, and regional intelligence systems. These resources allow officers to perform a wide array of investigative tasks. Officers routinely use them to identify wanted persons, locate stolen property, review criminal histories, and share intelligence with other agencies.

An ALPR database performs a very similar function. At its most basic level, it records that a publicly displayed license plate, attached to a particular vehicle, passed a particular location at a particular time. ...

Thus, consulting an ALPR database is not fundamentally different from querying a license plate or driver's license during a traffic stop or checking a suspect for warrants. Each task involves access to information for a legitimate law enforcement purpose. ...

Sure. Pull the other one. It jingles. These things are obviously not the same. One usually involves a traffic stop or, at least, human effort. That requires either a reason to engage in the query (the traffic stop) or enough desire to manually collect and run license plate numbers. Throwing out acronyms and drawing false equivalencies doesn't help the argument, but it does help confuse the issue - and so the tactic is useful.

Much later in the article comes the acknowledgment that these things are in fact different, but with an assertion that the difference should be ignored and can be obviated through arbitrary distinctions:

A routine license plate check begins with a vehicle that is already encountered by an officer. By contrast, ALPR cameras automatically collect the plate information of all vehicles that pass by. A network of cameras can then assemble those observations much more efficiently than individual officers watching traffic.

One concern that has been expressed is that ALPRs can be used to reconstruct one's location and activity history. This is a legitimate concern, and the solution is to adopt rules suited to the scale to which an ALPR system is used. Although a single ALPR hit is not the same thing as GPS tracking, broad or prolonged tracking should require a higher level of approval. In some cases, warrants should be required. However, short-term searches for a stolen car are not the same as reconstructing months of someone's movements.

In this case, they are the same. The dragnet is there and the panopticon is built. Setting a time limit on queries that needs a rubber stamp to override it is not a solution. To think it could be a solution is to still be burdened by institutional trust. That the writer has this burden is obvious from elsewhere in the article:

Every useful technology comes with risks. An officer can misuse a driver's license database to obtain an ex-girlfriend's address. A dispatcher can access criminal history or CAD notes without a legitimate reason. A detective can leak confidential information to a friend. Such conduct is already prohibited, and many officers have lost their law enforcement certifications and have been prosecuted for improperly accessing restricted information.

The problem with the "misuse" objection is that it proves way too much. If possible or actual misuse is enough to stop the public implementation of some technology, then it would apply to almost every government tool or power that has ever been used (whether in a criminal, civil, or administrative context). ...

Your terms are acceptable. Getting rid of ALPR systems would be a good start on a wider stand-down of mass surveillance technologies. The more the merrier. None of it - and none of its operators - can be trusted. "Misuse" (which I increasingly think is the wrong word, since these systems appear to be designed to be used in the "wrong" way) is not a risk, it is an inevitability - and it occurs almost immediately.

The entire class of thinkpieces from "conservatives" trying to rationalize mass surveillance technology with a rich history of egregious behavior by its operators is baffling. What about mass surveillance is suddenly rock-ribbed? Nothing. It comes from, I think, two sources. The first is vestigial trust, the belief that the system can be "safeguarded against abuse." It can't. Recent and more distant history does not show any success in safeguarding any system against abuse, which is why good systems design aims for systems that are smaller rather than larger and decentralized rather than centralized.

Outside of a poorly-construct and weak appeal to non-existent trust, I think there may be something more pernicious involved in some this: short-sighted faith that it's okay because "we're in charge." If you didn't click the links, go read the EFF and ILJ reports I referenced just below the fold. EFF in particular chose to cherry-pick specific abuses that are mostly around communist agitators like "No Kings" and abortion clinics while the ILJ focuses more on immigration and personal abuses like stalking, with no mention of any of these dragnets going against conservatives. Reports like these help provide a false sense of security to the "other side" - these systems are working for you! Comfort in advocating for these systems simply because "your side is in charge," if that is part of it, is short-sighted and foolish to the point of stupidity because even if it's true - and it isn't, as the last decade thoroughly demonstrates - "your side" will not be in charge forever.

Articles like this one that extol putative benefits but not the record of abuses, choose "safeguarding" and "rules" as a solution, hearken back to a world of trust that no longer exists, set up strawmen and drift into "the ends justify the means" territory on the back of bloody pathos are themselves prime evidence of core weakness of argument. Universal mass surveillance dragnets are a bad idea, are designed to be "misused" and are antithetical to liberty - and mental gymnastics in weak thinkpieces can do nothing to change that