A good piece in The Washington Post about an anonymous young American hacker is coming under scrutiny (thanks, Techdirt) for possibly revealing, inadvertently, the location of the hacker. This may have been done by what is called metadata, or extra information, in the photograph that accompanied the article. The hacker’s face in the photo is only partially in view, but some individuals have studied the metadata which seems to identify the photograph as having been taken in small town in Oklahoma.
Using other information in the article, which gives a pretty complete description of the hacker — “tall and lanky, with hair that falls down to his eyebrows… lives with his folks in a small town in Middle America .. nearest businesses are a used-car lot, a gas station / convenience store and a strip club…” some other commenters have tried to narrow down his home even further. Armed with such apparent detail, has The Washington Post inadvertently blown the cover of someone who asked them not to?
Actually, it’s not yet clear. The author has made it clear he is aware of the comments, and, according to one article, said he doesn’t want to comment about speculation. Someone claiming to be the hacker in question says on a discussion on the Washington Post reporter’s blog, that his location according to the metadata is “way off from where i reside”, adding that according to the reporter “it was old metadata”. The picture itself has been withdrawn, raising another commenter on the same blog to ask: “If those photos had old metadata, then why were they removed from the WP article and from the WP servers? It’s suspicious at best, and looks like destruction of evidence and interference with an official investigation at worst.”
I don’t know enough about how these kind of fields are entered into cameras to be able to offer any useful comment on this. It would appear that the fields in question are from the International Press Telecommunications Council’s list of NewsCodes, which sets a standard of fields for news organisations for better retrieval and database compatibility. So should the fields be trustworthy? I assume unless the data is automatically generated — technical information such as device model, focal length, exposure time etc, or location data by a GPS unit linked to the camera, say — the metadata itself should not be relied on too heavily, in or outside court. If the data is entered into the camera manually by the photograher it’s hard to imagine that being done conscientiously with each location change; if it’s done when the photos are moved to a laptop or desktop, then perhaps it might be more reliable but still could hardly be considered firm evidence.
Whichever way around, it raises one or two interesting questions. If the data is accurate, what other news photographs might throw up interesting secrets if we looked hard enough? If the data wasn’t accurate, what does that tell us about the usefulness of NewsCodes and other such metadata?