Surely anyone reading this post remembers with sadness and dismay the Boston Marathon bombing: two bombs exploded on April 15, 2013 at the finish line of the famous marathon (it’s part of the World Marathon Majors) causing three dead and over 150 injured.
After the attack, the FBI and the local authorities began the corresponding investigation. After three days of intense work, on 18 April, the FBI released in a press conference photos of two suspects taken from surveillance cameras. The agent in charge, Richard DesLauriers said literally: “Someone out there knows these people.”
What happened next is that agent DesLauriers launched a crowdsourcing process that shows how well and how bad this kind of processes can work.
When crowdsourcing works…
To better identify the suspects, the FBI used crowdsourcing to get images and videos about what happened there, a technique (the crowdsourcing) already used in the past by the governmental agency (i.e. for the analysis of written codes).
The FBI agent DesLauries asked citizen and media participation in order to capture the two suspects. He invited people to send any photo, video or material obtained during the attack and that could help in locating and identifying the two terrorists. About the identification, the agent insisted that only should be used for this purpose the photos provided officially.
The success was overwhelming. Thousands of images and videos were sent to the FBI, which even enabled an online form for easy uploading the material. In this case the crowdsourcing served its purpose, providing the authorities the vision of the attack from points of view better than those given by a surveillance camera, and allowing the correct identification of the terrorists.
When crowdsourcing fails…
Parallel to the police investigation began a process that ended becoming one digital witch hunt.
Despite the request of the agent DesLauriers to only use the official photos for identification purpose, many users began to upload their own pictures to Internet. Using web sites like 4Chan or Reddit, many people began to analyze and comment the pictures trying to identify the two suspects by their own.
In this way, the crowd began to tag as terrorists to people who had nothing to do with the attack. These people just had some ethnic trait, clothing or backpack similar to that used by some of the terrorists.
Such was the case of a young student named Salah Barhoun, of only 17 years. This guy started receiving Facebook messages in which he was asked why he had committed the attack (among other niceties that surely people gave him). The fact is that Salah found on one of these web sites a picture of him tagged as a terrorist. He quickly had to go to the police to clear his name.
One of the most surprising things (by its gravity) was that some media, like the New York Post, echoed these pictures and even published the pictures on the cover. It is easy to imagine the consequences for Salah Barhoun.
Continues (and concludes) in the next post…