Business

Lessons from crowdsourcing the Boston bombing investigation

All it took was a couple of hours for high-school sophomore Salah Barhoum to have his entire world turned upside down. Up until the New York Post featured him on their cover as a person of interest in the Boston Marathon bombing, he was best known for being a standout athlete. But suddenly, through no fault of his own, he was being followed by strange men convinced that he was responsible for the heinous bombings at the Boston Marathon that happened just days earlier.  The FBI had not named any suspects yet, but the New York Post hadn’t got its information from the FBI; it had picked up Barhoum’s name from anonymous online commenters.

Barhoum and well over a dozen others were victims of shoddy online detective work. Their identities were broadcast publicly, and they were accused of crimes they had nothing to do with and maligned by the national media as terrorists. In reality, they weren’t even being investigated by the authorities involved with the case. Instead, the’d been commenters on Reddit and 4Chan, who believed they were guilty based upon their clothes and appearance. What started as an atypical request from the FBI to the public, asking people to send in any photos and videos they had of the bomb scene, quickly morphed into an ugly digital witch hunt; one where the crowd’s fears, prejudices, and suspicions were given credence, while guilt and innocence were doled out based on shreds of circumstantial evidence.

In the four days, three hours, and nine minutes between the detonation of the first bomb and the Boston Police Department tweeting that the final suspect had been captured, a new approach for conducting crowdsourced investigations was established.

Although media outlets have been quick to lump all of the crowdsourced efforts together, there were two very different processes occurring, which proved to have drastically different outcomes: Crowdsourced intelligence gathering — a massive success — and crowdsourced crime solving — an abysmal failure. The FBI only ever asked for the first, but both happened simultaneously. They each offer important glimpses into major issues surrounding the future of law enforcement, justice, and surveillance.

In many ways, the Boston Marathon provided one of the most compelling cases for crowd involvement, ever. It’s one of the largest athletic events in the world (event planners estimate upwards of 500,000 people attend each year); the vast majority of attendees have smartphones, and a sizeable portion of those were actively taking pictures and videos throughout the event. Surveillance cameras have become ubiquitous, but they are fixed in place and have large blind spots. People, on the other hand, can provide deep context and multiple points of view of the same situation. For that reason, it’s a natural fit for Big Brother to look to tens of thousands of “Little Brothers” for help in gathering intelligence. After all, there is no police snooping network that could rival the surveillance regime of our smartphone lifestyles.

But regardless what the FBI wants, it can’t stop people from trying to conduct their own investigations. Events now play out in real time. The ability for a person sitting at home to have access to rich, detailed information about an event, as it happens, is magnitudes greater than what was available in the past. It is unlikely that this trend will reverse. Human beings love to speculate and gossip; now, we just do it together publicly. We’ve been empowered with communication and collaborative tools far more powerful than our own understanding of them.

When Richard Jewell, the security guard/hero of the 1996 Olympic Bombings, had his life and reputation destroyed by false accusations from the media, he was able to sue them for compensation. But when Sunil Tripathi, a student at Brown University who has been missing for the last month, was declared to be a vicious murderer because somebody misheard a police scanner, all his grieving family received was an anonymous apology. One of the largest questions these events raise involves defining what free speech online is. Is a person allowed to make a deeply damning accusation about somebody based on suspicion if others are doing it also and they believe it to be in the public interest?

The traditional media could play an enormously valuable role here by separating fact from fiction and providing verified, trustworthy information. Instead, most outlets just repeated false claims made online — providing a megaphone to statements that should never have seen the light of day.

The opening of investigations to the public is going to happen, and law enforcement groups have no choice but to embrace and try to get ahead of this trend. Even in this investigation, the crowd’s own efforts forced the FBI to release information earlier than it had planned. The FBI made the wise decision to set up a page where people could upload potential evidence, but before it even got around to creating this, citizens had already set up their own page. The medium on which this case was discussed made a difference in the quality of discourse as well; for example, Reddit fared better than 4Chan because of a system where true information could be “upvoted” and false information could be “downvoted.” Where the conversation about the next case takes place will directly shape how it plays out.

Despite its failings in this case, the crowd has proved itself to be an important force for public safety in the past. Like with many crowdsourcing-related activities, individuals are good at providing information or reporting events, but it is the next stage — taking action — where things often fall apart. The more passive their role, the more effective they have been. Seattle’s Police Department runs a program where citizens can receive tweets about and report when they spot stolen cars. German police have experimented with posting sketches of wanted criminals on Facebook, where citizens’ identifications have already led to several arrests. In another example, a Broward County Sheriff has leveraged his 10,000 Facebook friends to successfully track down stolen goods.

We now live in a world where information moves faster than we can assess its value… this is especially true in times of panic, disaster, and crisis. While an active terrorist investigation might not be the best place to allow wannabe detectives with no training, there are certainly situations where they can be helpful. We each have the potential to play a role in being guardians of public safety, but this requires from us a large degree of focus, caution, and care. Ultimately, these are the things that separate an empowered crowd from a raging mob.

Tarun Wadhwa is a research associate at Singularity University researching how advancing technologies can be used to solve public policy issues. 

Follow him on Twitter @twadhwa


We're studying conversion rate optimization. Take our quick survey and we'll share the results with you.