What is PRISM?
PRISM is a large scale, government-run, metadata surveillance platform. In other words, it’s a complex computer program built to gather, evaluate and communicate automatically generated personal information. It was created and supervised by people working for the United States National Security Agency.
PRISM’s functions are to collect, store, sort and generate reports on how people act on the Internet. This is also called behavioral tracking. Behaviors of interest might include things like which words people search for, which sites or pages people visit, or the contents of an email. They can include what time of day that person shops, what time of day she switches from focusing on her work computer to her cell phone, or where she keeps her money.
Often, behavioral tracking uses lots and lots of actual data points to extrapolate useful averages. One result of this is that the data points become effectively anonymous, because the research focus is on trends and not individuals. An individual would only stand out if her behavior was both highly unusual and in some way relevant to the goals written into the computer program. (Some individuals think they must be of interest to programs like PRISM, because they occasionally Google the phrase, “Ralph Nader 2016.” These people are perhaps clinically interesting, but probably not otherwise of interest.)
Here’s an example of how behavioral tracking works. We might ask our system to calculate when people in Albuquerque check their email for the first time in the morning, on a series of Mondays. Then we could have the program crunch the numbers and tell us that, on average, Burqueños between the ages of 18 and 25 check their email around 9am, while those between the ages of 45 and 65 check it closer to 7am. (I made that up.) We could use that information, combined with physical location, to send the digitally connected citizens of Duke City a coupon for a free cup of coffee at the Flying Star nearest them, when they’re most likely to notice and use it, before it gets buried by other email.
Why would the NSA want you to have free coffee? Likely it would not. But that is what the Internet’s behavioral tracking infrastructure was created to do, because it was not originally created for the NSA.
What’s the story with behavioral tracking?
At its most basic, behavioral tracking involves observing and recording people’s actions over time, usually to test a theory or to collect data that might aid future decision-making about that person or group. It has its modern origins in anthropology and psychology, but it probably started to most profoundly influence our society when seminal behavioral psychologist John B. Watson began working in the for-profit advertising sector in 1921, for a yearly salary of $10,000. As of 2013, the Wall Street Journal reports that behavioral tracking is worth $2 billion annually to businesses in the United States alone.
Most advertising in most media - television, print journalism, even outdoor billboards - uses behavioral data to match products and messages with audiences that are likeliest to “convert” - an eerily religious term that advertisers use to mean “buy something.” But the Internet offers a brave new world to behavioral trackers partly because it’s in the nature of computers to save everything, to record every nuance of every operation. In addition, online interactions are often much more personal and personalized than exchanges with other media (when was the last time you told your TV which year you were born, or asked it for more information on “genital warts vs. razor burn”?). As a result, many basic structural elements of the Internet are direct outcomes of behavioral tracking; the news stories we see on sites like Yahoo! and the text ads we see alongside search results on Google both result from those companies’ well-informed algorithmic estimates of what we’ll find compelling, and they’re often correct. If they weren’t correct, then our behaviors wouldn’t be worth the marketing investment, and we wouldn’t get to enjoy those free stories and search results in the first place. They’re not free; they’re paid for by advertisers who want to monetize us, and who can do it because they know what we like and what we fear, thanks to behavioral tracking.
I mention all this because I think it’s important to understand that PRISM wouldn’t be possible if, as a culture, we didn’t rely upon purportedly free services that we actually purchase with very high degrees of personal disclosure and capitalistic codependence. (Disclaimer: my capitalistic codependence is particularly complex. Between 2009 and 2012, the years Edward Snowden was contracting for the NSA, I contracted for Yahoo! and Google. Both of these companies have been named as contributors to PRISM’s database.)
Further, it’s important to understand that behavioral tracking in the age of modern advertising has evolved into a highly sophisticated and generally obfuscated instrument of persuasion. It’s this instrument, and not simply an observational technology, that our government now wields in the form of PRISM.
What does PRISM stand for?
It stands for Planning tool for Resource Integration, Synchronization, and Management. What this means is, roughly, “let’s use computers to tell us which people and groups are most likely of interest to us, so we can figure out where to spend money and assign staff.” Would an English teacher perhaps point out that integration and synchronization are very nearly the same thing? She would - this one did - and she might after looking it up realize that in this context integration means blending data points (a.k.a. “merging and deduplication”) while synchronization means matching them to corresponding data points from other sources. So, “let’s use computers to consolidate what we all know, make sure we all know the same stuff at the same time, and help us figure out where to spend money and assign staff.” That’s PRISM.
Ok, but what’s a real prism, and is the acronym a good fit?
A prism is an optical instrument - a piece of sculpted glass or something like it - that refracts light. That means it separates apparently white light into bands of different colors. All the colors were there the whole time, but the prism displays them in a way that’s visible to the human eye. In PRISM, data that was there the whole time is displayed in a way that’s visible to NSA operatives. So, yes, it’s an apt acronym. Meanwhile, “United States Electronic Review Protocol,” while perhaps even more apt, would yield the more unnerving USERP.
What can I do about PRISM?
In general, using the Internet means sharing some degree of personal data with third parties who hope to benefit from it. Whether that’s a fair trade-off is largely a matter of personal judgment, but essentially this is just the newest draft of a pretty standard social contract. In response I don’t recommend tinfoil hats, overly aggressive privacy software, or anything as crazy as Googling “Ralph Nader 2016.” Instead I suggest being aware of the supply chain that leads to your favorite infotainment and search results, and thinking critically before you click.
(NSA eagle by Electronic Frontier Foundation)
November 04, 2013