Indiscriminate surveillance at MIT

Logging of card access data at CSAIL has no clear benefits

By Richard Stallman Sep. 12, 2014

In the Stata Center, the doors to enter the building and then to enter the lab areas are opened by RFID cards (I call them “pox cards”) instead of metal keys. When the Computer Science and Artificial Intelligence Lab (CSAIL) moved into Stata in 2004, the MIT administration decided, without consulting the personnel of CSAIL, to make the system log which cards are used to open which doors and when.

At first, many of us were incensed about this; I remain so. A committee was launched to study the issue, but it dragged on, petered out, and never made a formal recommendation.

Thanks to Edward Snowden’s heroism, we know that digital systems accumulate massive records about everyone, and these are easily available to the state. Digital technology makes it easy to record events, and standard engineering and business practice is to log all data.

If we don’t want a society in which everyone’s movements and communications are recorded for Big Brother to scrutinize — if we want democracy and journalism to be safe — we need to make systems stop creating dossiers about us. (I’ve written about these general points before in gnu.org/philosophy/surveillance-vs-democracy.html).

MIT’s lock logging is one of these systems, and I felt I should take action once again locally, as well as globally.

When we criticize MIT’s practice of logging people’s movements, the Card Office responds with two arguments. First, they say MIT’s policy for access to these data is very strict.

Perhaps we could trust MIT on this if this were really under MIT’s control, but the USA PATRIOT Act allows State Security forces to collect all of this data at any time. MIT can’t refuse requests for data, or even tell us how often they happen. The NSA, whose unofficial motto is “collect everything,” may collect the all the data regularly, as it does with the major US phone companies.

Second, they assert that logging helps fight crime.

Does it really? Such claims must be put to the test. The NSA claimed that surveilling everyone in the US was vital for preventing terrorism. When it had to give details, it became clear that the supposed benefit did not exist.

MIT has not even tried to tell us what good it has achieved by surveilling people who work in Stata. I decided to ask. I emailed MIT Police Chief John DiFava to ask for information on whatever tangible good the Stata Center lock logging has done: for instance, criminals that have been caught, and/or property recovered, with the help of these logs. I emphasized that I sought no personal information, only aggregate data.

In his response, however, he did not provide the data, instead refusing to even acknowledge my specific requests. His message contained only general platitudes. I would post it, but DiFava didn’t respond when I asked for permission to publish his message.

If these logs are indeed protecting us from crime, surely the MIT Police would be glad to demonstrate just how effective they are. I call on them to show us with real data how much good MIT does for us through the logging of our movements. If they don’t, I think we can guess why.

Dr. Richard Stallman is the founder of the Free Software Foundation and a free software activist, as well as a visiting scientist at CSAIL.

Tags
guest column

6 Comments

Anonymous over 9 years ago

We should reserve alarmism for things that are actually alarming. This is an absolutely tiny amount of data about something completely inconsequential. I'm in favor of reducing the NSA's power, but crying wolf about every piece of data that gets stored just dilutes the issue and makes the cause seem ridiculous.

"Second, they assert that logging helps fight crime. Does it really?"

If there were a theft, it would help create a list of suspects. If there were a string of trespassing incidents, it would help figure out where trespassers are tailing people in or whatever.

Anonymous over 9 years ago

I am somewhat in agreement with your first statement, but, while there are certainly more important issues that we should be concerning ourselves with, that doesn't mean that we shouldn't still take the simple step of disabling logging of the card readers. It also sets a boundary to make further encroachment on privacy by the administration less likely.

I also wish to point out that since the card readers don't record when you leave, I am doubtful that the logs could be of any use in creating a list of suspects. Determining the most common access points to a building, which I think is what you are suggesting as a use in your final sentence, requires only anonymous counting of accesses. I cannot think of any use of personally identifiable card access data that would be admissible in court.

While Richard restricts his comments to the card readers, I have noticed a general rise over the past few years of cameras and other boondoggles around campus. I suspect that this is a result of bored administrators who are allocated too much money, which is then wasted on causes which do not contribute in any way to MIT's mission. MIT needs fewer administrators, and especially fewer of the type who think they are entitled to funding and importance despite their lack of valuable contributions.

David Andersen over 9 years ago

Perhaps Chief Di Fava could provide one statistic: The number of thefts at Stata (or campus-wide, in buildings that are open access during the daytime) that have been attributed to people with legitimate access to the building vs. the number to people whose access would not have been logged in the first place. It might be illuminating -- or disappointing -- either direction.

To Anon#1, remember that there is a cost to retaining information, and there is an even higher cost to retaining private information. A question to ask in this context, ignoring the privacy implications, is whether this is actually the best use of institute resources compared to other options for spending the same money.

Don't blindly take on faith that such measures do help crime. Consider the results of the increase in cameras in London; numerous studies have shown that the results are that they move crime more than they reduce it. Intuition usually benefits from being double-checked by real data.

Cruftily-alum yours,

Anonymous over 9 years ago

Back in 2006 when I started my undergrad at Marquette University, we had magstrip key cards for everything. Admittedly not as high tech as RFID, but with all the same tracking abilities.

Entering your dorm required handing your card to a human attendant to swipe, to verify you lived there. Of course, it could be easily bypassed by sneaking in when a large group entered the dorm, so it provided no real security, only tracking. I was once questioned about a crime by "public safety" (a private security company owned by campus, with a reputation for impersonating police), but later found out that I was actually above suspicion because my card swipe logs gave me a full alibi.

All the campus buildings locked after about 6pm, and required a card swipe to open the doors. The authorization database was always out of date, so I was never able to access the electronics labs that I needed to get to, but the attempt was certainly logged. I frequently called public safety to try to get them to open the door for me, and they said that they saw my attempt, but couldn't help me because they couldn't verify that I was allowed into the building. Eventually, I just learned to unscrew the bottom of the gray "software house" box, and short two wires together to unlock the doors. This solved both the authorization problem, and the tracking problem.

Anonymous over 9 years ago

I'm more concerned about the lack of public access to Stata (and every other new MIT building) outside business hours. One of the traditional good things about MIT has been the open-campus policy, and students and staff shouldn't let this slip. (The door at the eastern end of the Infinite doesn't even have a lock.)

A few years back at Harvard, there was a story in the campus paper about police barging into a sleeping student's room to interrogate him about a theft, since he was the last person to swipe in before it happened. I'll let y'all decide if this is a good thing which MIT should emulate.

Anonymous over 9 years ago

5: The problem in your story is overzealous police who think that such small evidence warrants a raid. We've had security cameras around for decades and few people object to those, and the same kind of thing can (and unfortunately does) happen from video footage, it just takes longer to ID the person.