Topic: Telecom, Internet & Information Policy

Sensors and Social Consequences

A “sensor” is a device that measures a physical quantity and converts it into a signal that can be read by an observer or instrument. Sensors that convert analog information into digital form are the most interesting. The information they collect is easy to store, transmit, and reuse.

Digital sensors are all around - the keyboard on your computer, your cell phone, the surveillance cameras in your office building, and so on.

Lots of good things come from having these sensors around, and the systems they attach to - that’s for sure. But they don’t always serve our interests. Let’s take a look at an example of digital sensing gone wrong.

A colleague of mine recently returned from a business trip, where he engaged in important and sober work. He arrived home late from his trip, and his patient and loving wife, already in bed, engaged him in some conversation. Fairly quickly, she asked him whether he had enjoyed himself at the strip bar (!). My hard-working and serious colleague was concerned. Why, on returning to the warm glow of his happy home-life, should he be asked this question?

As he tells it, he found himself short on cash one evening, and ducked into the nearest establishment looking for an ATM. The generous purveyors of this … nightclub - who could have known it was something more? - graciously allowed him entry for the few moments it took to get the cash and be on his way.

ATMs are digital sensors. They record usage information and tie it to other details, like location. This is known as “meta-data” - information about information, such as where and when a given piece of information was collected.

The ATM transmitted this data and meta-data back to my colleague’s bank and, through an online banking system, to his wife. The system identified the ATM as being at “Antics Topless Lounge” or something like that. You can understand the short string of inferences that his caring, truly lovely wife drew when presented with this single item of sensed data.

The reporting of ATM location information is a convenience to those who may have forgotten where they used the ATM, but it’s less welcome to someone whose sweet and lovely life-partner might draw unfortunate inferences from ATM use in certain locations. Sensors have social consequences, and they’re not all good.

So I was nonplussed by the cover of the latest issue of Government Technology magazine. It shows the front of a police car, photographed from a low angle to give it a pugnacious look. (Alas, I can’t find the image online.) The car is decked out with lights and sirens, of course, but also with sensors - on the roof and behind the windshield.

“FREEZE FRAME,” says the magazine cover, “License plate scanners extend the reach of local police.” Inside, an article describes how license plate scanning by U.S. police agencies is “the next big thing” for catching stolen cars and locating suspects. But the real benefit, according to the chief of detectives and commanding officer of the Detective Bureau at the Los Angeles Police Department, “comes from the long-term value of being able to track vehicles - where they’ve been and what they’ve been doing … .”

Make no mistake: there is value in that, just like there’s value in knowing where you used the ATM. But there’s risk in that, too. It’s not an unalloyed good to give people data about your comings and goings - other than your loving, caring family, of course.

Unlike my colleague and his saintly wife, it’s none of the police’s business where law-abiding citizens have been going and what they’ve been doing. When these sensors are used for mass surveillance and not just spotting bad guys, that crosses an important line.

This is not an argument against giving police these sensors. They will be a boon for law enforcement and an aid to our safety and security. But if the back-end systems put information about every vehicle’s location into a database for later use, that’s inappropriate surveillance of the law-abiding public. Unlike my colleague’s charming, gracious, and forgiving wife, the police shouldn’t be in a position to ask us whether we enjoyed ourselves at the strip bar.

Abstract Ideas Can’t Be Patented. Or Can They?

The Supreme Court has long held that laws of nature, physical phenomena, and abstract ideas are not eligible for patent protection. Because these things are discovered rather than invented, they are “free to all men and reserved exclusively to none.” In recent years, however, the United States Court of Appeals for the Federal Circuit, which hears most patent appeals, has begun to relax the restriction on such patents. I’ve written before about the problems created by software patents. Software is is ultimately just a sequence of mathematical formulas, and in their pure form they’re not patentable. But in a series of decisions in the 1990s, the Federal Circuit opened the door to patents that cover software when it’s loaded onto a computer, which of course is the only useful thing to do with software. Since then, we’ve seen an avalanche of patents on software, which have started creating serious problems for innovators in the software industry.

The latest example of the problems on patenting abstract concepts comes via Mike Masnick of Techdirt: a company had some problems with a satellite launch, and wanted to use a maneuver called a Lunar flyby to correct it. Unfortunately, Boeing holds a patent covering the maneuver they wanted to use, and they have been unable to negotiate a license of that patent. So they’re planning to let the satellite go down in flames and try to collect the insurance money on it.

Now, as Mike points out, the maneuver in question is just an application of basic physics to spaceflight. The basic principles have been understood since Newton, and NASA has been computing these kinds of orbital trajectories since the 1960s. The patent office should have rejected the patent for trying to patent a straightforward application of basic physics. Unfortunately, thanks to the Federal Circuit’s increasingly permissive standards for patentable subject matter, Boeing was granted the patent, and this company now faces the unappetizing choice of leaving the satellite in the wrong orbit or getting embroiled in litigation with Boeing.

Crucially, the Supreme Court has never endorsed the Federal Circuit’s experiment with allowing patents on abstract ideas, and several justices have voiced concerns about the direction the Federal Circuit has taken the rules for patentability. Apparently, the widespread outrage over the abuse of such patents has gotten the Federal Circuit’s attention, as it has decided to re-hear a case called In Re Bilski that could give it an opportunity to tighten up the rules for patenting abstract concepts. Several public interest groups have filed briefs in the case urging the court to do just that.

The Federal Circuit will be hearing the case en banc next month, and it has already become one of the most closely-watched cases on the Federal Circuit’s docket. Given the Supreme Court’s heightened interest in patent issues in recent years, it’s not hard to imagine the Supreme Court deciding to review the decision as well. Given that Congress has so far ducked the issue of reining in patents on abstract concepts in its pending patent reform legislation, In Re Bilski may be our best chance of reform.

The Helping Hand of Government …

… strips away privacy before it goes to work.

Here’s a nice, discrete example: S. 2485, introduced in the U.S. Senate last week, would require asset verification of participants in State Medicaid programs, exposing the personal information held by financial institutions to government access.

This privacy loss is a natural outgrowth of entitlement programs. It’s nearly mandated by the simple and warranted effort to reduce waste, fraud, and abuse.

My 2004 Policy Analysis, “Understanding Privacy - and the Real Threats To It,” explored how entitlement programs almost always carry with them a significant privacy-cost:

To provide benefits and entitlements—and, of course, to tax—governments take personal information from citizens by the bushel. Nearly every new policy or program justifies new or expanded databases of information—and a shrunken sphere of personal privacy.

The Vote: Ease? Security? Or Enough Already?

The Universal Right to Vote by Mail Act of 2007 (H.R. 281) recently passed the House Committee on House Administration. It would amend the Help America Vote Act of 2002 to require states to allow eligible voters to request a mail-in ballot for all federal elections without having to provide a reason.

In a TechKnowledge piece called “Voter ID: A Tempest in a Teapot that Could Burn Us All,” I shared some thoughts that are relevant to this bill:

Increasing voter participation has been a policy fetish for the last decade or two-never mind whether more voting for its own sake makes a better democracy… . The growth in absentee balloting has undone some of the protections against voter impersonation and multiple voting that previously existed. People are much more reticent to commit fraud in person - it’s riskier - so in-person voting was a natural security against impersonation fraud. Voting in multiple jurisdictions is simply too time-consuming to do on any scale when it has to be done in person.

The bill would require states to verify signatures on absentee ballots by cross-checking them with voters’ signatures on the official list of registered voters, but this only begins to shore up the security hole opened by mass absentee balloting.

The people who want this bill undoubtedly believe it will improve both the political discourse and their electoral prospects. Folks on the other side - the proponents of identification requirements for voting - will only be energized by these efforts, which lower the bar for both legitimate voting and for voter fraud.

Both sides should just drop this food-fight-to-the-death and work on substantive policies that they believe will win voters to their sides. Hopefully, those policies are centered on limited government, free markets, and peace.

“Biggest … Lie … Ever”

A friend and supporter of my work on REAL ID sent me a link to this WebMemo from the Heritage Foundation, entitled “All Aboard: Fifty States Now Compliant with Real ID.” I’m using the subject line of his email as the title of this post.

There certainly seems to be confusion in some quarters about REAL ID’s current status. Let’s take a brief look at how states stand in terms of compliance.

Because not a single state will comply with REAL ID on the statutory deadline, May 11th, the Department of Homeland Security has been giving out deadline extensions willy-nilly the last few months. It gave extensions just for the asking to states that have statutorily barred themselves from complying, for example.

Some states refused to even ask for extensions. When this happened, DHS quickly switched to issuing states extensions if the states were independently changing their driver’s licensing processes in ways that would meet any of the requirements of REAL ID. States like Montana and New Hampshire wrote to DHS expressing no intention to comply with the law, but stating what they had done on their own. These DHS interpreted as requests for extensions, and granted them.

When the governor of Maine last week finally sent DHS a letter stating his intention to submit legislation relating to REAL ID compliance, the DHS took that as a request for an extension and granted it. The Maine legislature will have to consider any such bills, of course. Maine’s is the legislature that was the first in the country to reject REAL ID.

Getting deadline extensions by hook and by crook out to all 50 states is a pretty long way from getting all 50 states to comply. The actual state of things is reflected well on this map, maintained at the ACLU-run Web site It shows seven states still self-barred from complying, and many others protesting the law. An eighth - Idaho - recently saw legislation barring compliance with REAL ID move through the Senate and to the governor’s desk.

Some Myth-Busting Is Quite Revealing

After DHS Secretary Chertoff’s testimony to the Senate Judiciary Committee this week (at which he was apparently rebuked for “bullying” states on REAL ID compliance) he sat down with a group of bloggers to discuss things.

Congratulations are due the Secretary for making himself available in an open forum like this, especially because it allows us some insight into his thinking. It makes more clear why he and his colleague Stewart Baker feel a need to engage in so much REAL ID “myth-busting.” Though I have assumed their comprehension of the problems with REAL ID, perhaps I have been mistaken, as Secretary Chertoff does not exhibit a good sense of information technology or the information economy.

Here’s the myth that Secretary Chertoff purports to bust:

I had someone say to me today, “Well, when you have these REAL ID licenses with a machine-readable zone … it’s gonna be used to track people. People can skim it. And they can steal it. And then they can use it to follow you around.” Now this is a fantasy. This is just not true.

The Secretary overstates the argument and so shades into attacking a straw man, but the context is conversational. So let’s look at what the real argument is, and then at the Secretary’s responses. I touched on the question of tracking in my testimony to the Senate Homeland Security and Governmental Affairs Committee:

There are machine-readable components like magnetic strips and bar codes on many licenses today. Their types, locations, designs, and the information they carry differs from state to state. For this reason, they are not used very often. If all identification cards and licenses were the same, there would be economies of scale in producing card readers, software, and databases to capture and use this information. Americans would inevitably be asked more and more often to produce a REAL ID card, and share the data from it, when they engaged in various governmental and commercial transactions.

In turn, others will capitalize on the information collected in state databases and harvested using REAL ID cards. Speaking to the Department of Homeland Security’s Data Privacy and Integrity Advisory Committee last week, Anne Collins, the Registrar of Motor Vehicles for the Commonwealth of Massachusetts said, “If you build it they will come.” Massed personal information will be an irresistible attraction to the Department of Homeland Security and many other governmental entities, who will dip into data about us for an endless variety of purposes.

This is not an argument that the currently proposed REAL ID license would be read surreptitiously, as might happen with an RFID-chipped card. (The Secretary says that REAL ID currently does not require RFID, but neglects to mention that the “Enhanced Driver’s License,” which satisfies REAL ID, has one.) The argument is that a great deal more data about us will be collected.

This will include “meta-data” - information about the collection of information, such as time, place, purpose, collecting entity, and so on. Combined identity data and meta-data form footprints about our comings and goings. These footprints, collected in interoperable databases, combine to form tracks.

Perhaps it’s a complicated argument, but it’s a coherent one: REAL ID would lead to tracking of law-abiding Americans.

Nothing the Secretary says conveys that he’s aware of meta-data or actual data collection processes. He says that the machine-readable zone (or MRZ) is “nothing more than the information on the face of the license. I already have a reader for the license - it’s called my eye - and I can read what’s on your license. So therefore there’s nothing I’m going to get out of the MRZ that I can’t get from the face of the license.”

Alas, even this isn’t quite true. The regulation prescribes certain minimum data elements for the MRZ, but doesn’t restrict the use of others, and it doesn’t require states to restrict the content of the MRZ to only what is on the face of the license. The MRZ could lead to tracking of people and their activities based on their race, for example, a data element many states currently include in their MRZs. Despite receiving comments concerned with this during the rulemaking process - oh, and in congressional testimony - DHS declined to prohibit including race in the MRZ of REAL IDs.

Card readers are not just little electric eyeballs. They record information in digital form. This means that identical copies of these records are easy to store, easy to compile, easy to transfer, and easy to reuse. Collecting information in digital form is materially different from collecting information in analog form. Most people who work with technology know that implicitly. To be credible on identification technology issues, one must know this and acknowledge its significance.

Finally, the Secretary says that the DHS is not going to create a lot of databases using REAL ID. That may be his intention, but he’s in office for about ten more months. And whether DHS creates them or not, databases of information harvested using REAL ID would likely be available to DHS.

It is very hard to design information technology systems that do not collect and retain information. The current secretary’s personal opinion about databases just isn’t good evidence of whether or not there will be databases of information about the comings and goings of law-abiding Americans. Chances are very good if REAL ID is implemented that there will be.