I was invited to be a panelist at this year’s Personal Democracy and Freedom (PDF) conference held here in New York City. The panel was titled, “Battling Big Brother” and the idea was to comment on the degree to which individuals may be caught up in collateral damage from government collection and mining of data for the purpose of national security. I great question, indeed!
I wanted to make a few comments on that panel, and thought I’d reproduce some of them for this blog below.
I’m sure by now everyone is familiar with the hype around collecting and mining big data for individual patterns. And it’s not going to shock anyone to state that government, just as with private sector (e.g. facebook and google) have great interest in doing this.
As far as commercial interests are concerned, from what I see, these often focus on advertising — how can content providers effectively identify their visitors in order to present them with relevant ads? On one hand, the consumer benefits are obvious. Think of all the free online services and mobile apps that we use every day — they are likely supported by advertising. On the other hand, there are privacy concerns when people are tracked, and other personal characteristics inferred, without their consent (e.g. target pregnancy girl). Moreover, there may be economic consequences from price discrimination which may also be seen as unfair. E.g. when those of higher income receive greater discounts than lower income people.
Public interests of big data include, among other things, law enforcement and national security. But they have an advantage that private sector doesn’t in their ability to link many more kinds of disparate data sources and make more important inferences. They can combine CCTVs, drones, and of course, data collected from the private sector like phone records, emails, search engines, and network traffic from ISPs. I think we can all agree that the benefits of preventing bombings, and cyber attacks using these big data sources are large. What is of debate is how state agencies go about that and what tradeoffs we are willing to accept (e.g. PRISM and Verzion phone metadata collection).
I now want to talk for a few minutes about two recent news stories that I think are relevant to this discussion. The first is this week’s supreme court decision to allow DNA collection at the time of arrest for a violent crime. Ostensibly, this is done to because of the strong force of recidivism: the notion that a criminal caught for one crime may have committed some other, unresolved crime. The novelty — and risk — is that DNA is thought to be a better detection mechanism than fingerprints because it’s more difficult to conceal one’s DNA at a crime scene. But again, consequences occur when we feel that the government is overstepping its authority — when they suddenly have access to data we don’t think they otherwise should. What interests me most about the ruling, however, is the question: does DNA collection really work? I think there is a legitimate issue of whether law enforcement is more effective when they can obtain this information. I think this is important because if many more criminals are caught who would otherwise not be, then it becomes a discussion of tradeoffs. However, if there is no measurable effect, then the policy seems strictly bad. Similar questions can — and probably should — be asked of other forms of government data collection and surveillance: unless there is clear evidence of the effectiveness, where is the justification?
The other story is one authorizing military commanders to engage in what’s called ‘active defense.’ i.e. to hit back at attackers who conduct cyber attacks on military systems. The benefits of this style of defense have been debated (at least) in the IT security community for many years, and it’s interesting to see acknowledgement of this kind of behavior by the military now. Perhaps this is due to reportedly dramatic increase in espionage from China. There have also been calls by private companies (e.g,. those victimized by loss of IP) to engage in the same kind of behavior. What is not clear, however, is what force of retaliation is suggested, and what kind of collateral damage may be caused by this.
Now, to the question of what can individuals do? On one hand there are a host of privacy enhancing technologies and practices that individuals can employ: when searching online, you can use duckduckgo; when looking to browse anonymously you can use TOR; when purchasing groceries, you can use someone else’s loyalty card number; you can choose not to register a DC metro card; etc, etc. This makes us very empowered as consumers. However, on the other hand, at some point, you *will* leave a digital trail. You will need to go outside (where you’re likely to be captured on CCTV); you will need to buy something with a credit card, or take out a loan (adding to your credit profile); make a call on your cell phone; or you will simply forget to use one of those PETs. And so I’m quite conflicted regarding the extent to which individuals really have any power to control their digital trails at all. To me, the persistence and ubiquity online tracking and surveillance as an unstoppable force and that while we may be able to redact some entries from the mountains of data files we leave, I don’t see any practical solution to avoiding creation of those files to begin with.
PDF Program: http://personaldemocracy.com/conferences/nyc/2013/program