MIT Media Lab's 'Immersion' project reveals the power of metadata: uses individuals' Google Gmail metadata to create accurate maps of personal relationships
MIT Media Lab's new project aims to show individuals exactly how much personal information their ‘metadata’ reveals about them. Dubbed Immersion, the tool has been under construction for a long time but its release is incredibly timely.
Following revelations regarding the capacity of the NSA and GCHQ to snoop on individuals’ data, government ministers have attempted to reassure individuals by saying that such programs are collecting “only” metadata. Project such as MIT’s Immersion aims to show us exactly how misguided these reassurances are.
In this context metadata means the information regarding the context of any particular communication. The metadata for a phone call would include the time of the call, its duration and when it took place. The metadata of an email exchange would be similar – including the time sent and the people involved, but not the subject line or email content.
Politicians have seized upon this distinction and held it up as proof of their innocence - claiming that snooping metadata is less intrusive. This interpretation has been objected to in the strongest terms by academics and experts, who say that the general public are simply unaware how much personal information metadata reveals about someone’s life.
Plugging your Gmail address into MIT’s Immersion allows the system to scrape your email account for its metadata, and produces a complex bubble map showing who you talk to, how much you talk to them, and what your relationships with your contacts are.
Speaking to the Boston Globe, one of the leads behind Immersion, César Hidalgo said: “It’s like the world is catching up to what a fringe group of academics was aware of in 2004 and 2005. “Nobody liked [thinking about metadata], and nobody cared about us, and they all thought that working with mobile phone records or e-mails was sort of a curiosity or a stupidity.”
Although the tool is currently unavailable due to heavy demand, internet activist and author Ethan Zuckerman has written a fascinating blog-post detailing his use of the tool:
“If you are a member of a secret organization planning overthrow of the government, you’ve probably already thought hard about what your metadata might reveal. But if you’re an average citizen with “nothing to hide”, it may be less obvious why your metadata may not be something you are comfortable sharing.”
As Zuckerman notes, one particulary interesting feature of metadata is its ability to cluster people into accurate groups – work colleagues, or university friends for example. This is achieved by noting who you send out the same email to and also logging which individuals are ‘bridges’ between different networks.
Zuckerman describes his own graph, and its social groupings, as “a reminder that one of the reasons we might be concerned about metadata is that it shows strong relationships, whether those relationships are widely known or are secret.”
Put in the context of the NSA and GCHQ revelations, these insights into the power of metadata is deeply troubling. Writing for The Guardian, John Naughton speculates on how this situation has arisen – questioning whether it’s due to “the naivete/ignorance of legislators who swallowed the spooks' line that metadata-hoovering was just an updating of older powers to access logs” or if the political climate post-9/11 made democratic governments “easy meat for bureaucratic empire-builders in the security establishment”. As Naughton himself concludes however, its not the context that is most worrying, but that such activities have been conducted in the dark for so long.