How Social Media Policing and Online Vigilantism May Increase Wrongful Convictions

The police are in your home. No, not quite literally, but almost. Just like the billion-plus people who log onto Facebook every day and the thousands of self-promoters who brag on Twitter about crimes they’ve committed, the cops have been flocking to social media for several years now. From the Wayne County Sheriff’s Office in Ohio to the New York City Police Department, they’ve been setting up social media accounts, all in a bid to communicate more effectively with the public and, ostensibly, to solve cases. From the perspective of the forces involved, this strategy has worked wonders, with a litany of people incriminating themselves via boastful Facebook posts, and the public obligingly responding to closed-circuit television footage with the names of suspects.

Yet despite the noticeable benefits to police departments of harnessing social media and big-data technology to transform thousands (if not millions) of people into unofficial police informants, there are numerous demonstrated and potential downsides to this change in police operations. Not only does it open the floodgates of official police channels to the slews of misinformation often associated with the dawn of the internet, but also it threatens to stimulate a growth in misguided internet vigilantism and increase wrongful convictions. If this happens on any considerable scale, then the use of new digital technologies in policing, far from strengthening the balance of justice in the world, may only weaken it further.

One of the Biggest Crime-Fighting Tools

For some departments, the use of these technologies is already prevalent. In San Jose, California, the police recently chalked a 75 percent reduction in burglaries largely up to their systematic employment of social media and technology. According to a press release, they’ve begun exploiting a program that “almost immediately” posts images and surveillance video on their public portals. What’s more, these media postings have apparently witnessed an emphatic public response, with tips flooding in on most cases, and with six suspects being identified for the last 10 cases they’ve publicized. Whether these were reliable tips and identifications wasn’t disclosed by the department, but for the moment, that’s neither here nor there.

Similar social media boosts to police operations have been reported elsewhere. In Midwest City, Oklahoma, the police testified to social media having a comparable effect on their performance. Chief of Police Brandon Clabes declared that videos placed on social media were “helping the department solve more crimes,” and have become “one of our biggest crime fighting tools [the department has] in this day and age.”

Likewise, a USA Today article from 2012 documented how more than 40 police departments across the United States had already turned to YouTube and other social sites, with the Philadelphia police stating that social media had helped them solve 85 cases between February 2011 and June 2012 (the date of the article’s publication). A couple of years later, the International Association of Chiefs of Police announced that 95 percent of police forces in the United States use social media in one capacity or another, and that 82.3 percent of the forces polled employ such media for the purposes of pursuing criminal investigations.

In other words, social media use is now a well-established component of day-to-day policing. In fact, an indication of just how well established these practices are, and just how co-opted in the fight against crime the users of social media have become, can be glimpsed if you search Twitter accounts for “police.” Here, you will see a list of almost all the major police departments and organizations in the English-speaking world: the Metropolitan Police (the UK), the Toronto Police (Canada), the Mumbai Police (India), the South African Police Service, New South Wales Police (Australia) and the Nigeria Police Force. These institutions and more are now visiting social media sites daily, tweeting about and at “wanted” persons, posting images of missing persons and sharing various public service announcements.

The “Soft” Police State

Quite apart from the practical defects and downsides to the technological turn in policing (more on these soon), there is something rather unsavory and askew as a matter of principle about enlisting social media users as honorary members of the police. Given that some 72 percent of the online US population use social media sites (and 62 percent of the entire adult US population use Facebook), this equates to quite a large network of dormant informers, potentially – and sometimes unwittingly – ratting on their shifty-looking neighbors.

These tactics are unsavory because they have grave implications for civil liberties. Because users of social media as a percentage of the total population are so considerable, and because social media are so ubiquitous, their incorporation into routine police operations has the sheer capacity to transform the nation into a “soft” police state, at least insofar as they and the police will enjoy near-constant access to each other. Within this hypothetical state, the police will be able to process and monitor the public’s online activity without leaving their headquarters, while the public will have a very immediate and effortless means of reporting any “suspicious” behavior or “useful” information to the authorities. If such a scenario were realized, then the police would become a constant, if very discreet, presence in our lives, able to watch over us and make it easier for us to watch over ourselves.

That said, whether all 72 percent of the online US population will become part-time snitches will ultimately be a matter of how effective the police are in encouraging them to make the most of the new channels opened up by the internet, and this in turn will be a matter of police resources and policy. Still, if enough of the 72 percent are interested, and if the police continue increasing their encroachment into social media, then we may end up entering some all-too-real parody of 1984 (if the National Security Agency hasn’t already brought us there).

The infiltration by the police of social media and thereby “the social domain” in the abstract already represents something of a violation of privacy, a violation of the private-social realm by the public-political. Facebook’s mission, for example, states that the social media site is dedicated to “people” and their ability to “stay connected with friends and family.” It says nothing about providing the authorities with a direct access line to such people, or with the means of tapping the networks and social spaces they’ve developed as a source of information on criminal activity (then again, it’s already common knowledge that Facebook has betrayed its own self-declared mission in other respects). Even so, this is exactly what Facebook and other similar sites are doing: allowing social networks to be built so that they can be penetrated instantly by the police and their various initiatives.

Such a mass-scale opening of social groups to the presence of police is an unprecedented development, and in its wake, it may arguably spur the proliferation of social media users who consider themselves part-ordinary civilian and part-vigilante. In Australia, for example, there was the case of a concerned mother who mistakenly thought a man was taking a photo of her children in a shopping mall when in fact he was simply taking a selfie next to a picture of Darth Vader. The mother then took a photo of him, posted it on Facebook and reported him to the police, having jumped to the unfounded conclusion that he was a pedophile.

Even more alarmingly, there’s the development of Facebook-based vigilante groups in countries ranging from the UK and Germany to Peru, where people are now encouraging each other to apprehend potential burglars and sexual harassers, often in increasingly violent ways.

There’s even the possibility that social-media-based vigilantism is itself breeding a more generalized, offhand culture of self-policing, through which people are being shamed and chided even for more personal peccadilloes. For example, in July 2015, there was the story of the two sisters who tweeted footage of a married woman sexting another man while sitting beside her husband at a baseball game. It’s probable that with the increased visibility and activity of police on social media, this kind of haphazard internet activism will only be encouraged and motivated further, resulting in a climate where a growing number of overzealous people are “policing” and harassing each other in the pursuit of likes and retweets.

Trigger-Happy Identifications and Wrongful Convictions

Aside from its potential effects on culture and wider society, there are various legal and technical issues with the increasing reliance of police agencies on social media. For one, there is the potential that, far from being reliable, the information they receive from the public is racked with inaccuracies and distortions. By opening themselves up to millions of users on Facebook and Twitter, police potentially open themselves up to a greater quantity of misinformation and speculation. Examples of such misleading noise abound when it comes to the internet and social media, as is revealed most starkly by the Boston Marathon bombing and the initial misidentification of the individuals responsible for the atrocity. There are many analogous episodes of people being wrongly labeled as murderers via social media. As a result, police need to expend extra resources to sift through an expanded mass of junk. This situation also raises the disturbing possibility that wrongful convictions may increase in parallel.

That an increase in wrongful convictions is likely is evinced by the fact that, according to the Innocence Project, 72 percent of wrongful convictions are the result of eyewitness misidentifications. As for these misidentifications, they generally occur because people are susceptible to having their opinions on who they saw and who is visible in evidential imagery influenced by intervening suggestions, such as in the well-documented 1984 rape of Jennifer Thompson, who wrongly identified an innocent man as her assailant after being shown photos of known criminals by the police. What this means is that, with the increase of social media reporting of crimes by the police and news outlets, eyewitnesses are likely to be similarly swayed by the “intervening suggestions” this reportage provides.

Such a case of victims being influenced by “intervening suggestions” happened in a trial considered by the Toronto-based Neuberger & Partners LLP, who noted that the victim’s identification of her assailant in court was tainted by “her viewing [the suspect’s] picture on Facebook a day or two after the robbery.” The same article also notes that a judge asked that less weight be given to an uncle’s identification of a young person as the perpetrator of an assault on his nephew, since this uncle had seen the suspect’s Facebook profile – replete with weapon and gang iconography – before making the identification.

In these two examples, the courts involved were worried that social media may have skewed the witnesses concerned toward false-positive testimony. Just as they were worried about this, so too should we be worried that the growing use of social media may skew the police toward a similar outcome. This implies that the risk of social media doesn’t simply reside in the likelihood of false identifications from the public, such as with the Australian model who was questioned by police after being identified via social media as the culprit of the 2015 Bangkok bombing. No, it also resides in how the police actively mine and search social media sites themselves, preemptively flagging up likely criminals and tarring their online social networks as potential co-conspirators.

In Fresno, California, this is tangible in how police use new software known as “Beware” to calculate the “threat score” of individuals. Depending on their “data points, including arrest reports, property records, commercial databases, deep Web searches and the [individual’s] social-media postings,” suspects and persons of interest are classified according to a traffic-light system (i.e. red, yellow and green), with red designating the greatest threat and green the lowest.

Such software might make things easier for the police when readying themselves for a dispatch, yet at the same time it flags up one of the more unwholesome ramifications of “social media policing.” That is, given the racism already embedded in the identification of “suspects,” it’s highly likely that “threat scores” and social media profiling will disproportionately target Black people and other people of color. If so, the “Beware” system and the expanded social media use it represents have the potential to exacerbate the marked racial disparities that already wrack the criminal legal system.

Also, as the American Civil Liberties Union has already asserted, the Fresno police’s broad-brush approach to persons of interest may result in the police arriving at a scene prepared to take some unnecessarily heavy-handed and unfair action. Even though its inner workings are a closely guarded secret, the Beware program is likely unable to distinguish between someone who posts genuinely criminal material on their social media accounts and someone who, for example, is critical of the police and their policies (e.g. Black Lives Matter). As such, its existence is one more indication of how the use of social media may actually end up lowering the quality of policing, rather than improving it.

The Loss of Innocence

As other commentators have observed, the police had been trawling through social media long before Beware, and for the most part, their use of Facebook and Twitter has been distinctly problematic. Often, they use it to piece together outlines of gangs, using the available networks of friends, followers and likes to deduce who might be criminally associated with known outlaws. The thing is, this method also lacks considerable nuance and context, as it disregards the possibility that “friending” a person who has committed a crime or “liking” a video of a crime, for instance, doesn’t necessarily mean you’re actually in league with that person or have perpetrated that crime. In certain high-profile cases, this kind of simplicity has led to false arrests and charges, such as with Jelani Henry, who in 2012 was charged with attempted murder after liking posts by a Harlem gang, which counted his brother as one of its members.

Henry wasn’t the only person to be arrested primarily for his or her online activity. In 2012, the New York City Police Department launched Operation Crew Cut, an initiative that based itself around the monitoring of social media activity, much of it coming from Black people. Since it began, numerous raids have been conducted by the department, with the most infamous being a June 2014 maneuver in Harlem that netted 103 arrests in connection with two homicides, in the process substantiating the fear that social media policing will disproportionately prejudice people of color. In the indictments authorizing these arrests, “Facebook” appears more than 300 times, and even though many or most of the arrested individuals may have carried some degree of guilt, the example of Jelani Henry strongly insinuates that some of the 103 may have been innocent. Indeed, a City University of New York law professor said just as much on the subject, stating that the police are now using social media to “hold 50 kids accountable” for a single shooting.

These cases show that, as with asking the public to help in identifying suspects and persons of interest, using social media may ramp up the scale and speed of police investigations, but at the cost of losing subtlety and precision. However, beyond jailing the occasional innocent, there are fears that the strategy of pre-labeling individuals as “gang members” or “threats” may play a role in the courtroom as well, replacing the presumption of innocence – one of the fundamental tenets of the criminal legal system – with the presumption of guilt.

This was discussed in a 2015 law paper written by researchers from the University of London in the UK, who argue that the concept of the suspect now contains a recognition of guilt. They recount a drunk-driving Twitter campaign conducted by Staffordshire Police in the UK, a campaign that publicly identified people as drunk drivers, despite the fact that these people had only been charged with (and not convicted for) driving under the influence. While it’s not as public, the practice of building databases and networks of people who’ve been charged with crimes has a similar effect. It tags people as “not ‘wholly innocent,'” smearing them in the eyes of the police before they even have the chance to appear in a court of law and clear their names.

Fighting Vigilantism With Vigilantism

The paper’s authors note that the use of the technology surrounding social media may deceive the police into thinking that their inquiries are correspondingly “scientific” and “objective.” Because they amass data from the likes of Facebook in a more-or-less systematic way, they may cultivate an overly confident and hubristic faith in the evidence this data provides.

Perhaps this presents one of the biggest dangers of them all: that the police will think technology automatically makes them infallible. With this misapprehension may come an increase in the kinds of errors and injustices outlined above, as well as in the inability to see the latter for what they truly are. The police may begin accepting an ever-greater quantity of spurious statements from the public, and they may begin arresting an ever-greater number of innocent people, all the while convinced that their gravitation toward social media and big data insures them against such mistakes. In some ways, the receipt of false information and the arrest of innocent people will be nothing new for them, just as the existence of vigilante subcultures and bias against people of color are also nothing new.

However, given the massive scale afforded by social media and the internet, these unfortunate phenomena may very well proliferate at increasing magnitudes, combining an increase in social media vigilantism with an increase in presumptions of guilt. Such a lethal combination would lead to a surge in wrongful arrests and convictions, ruining lives and further eroding trust in the police at the same time. Ultimately, this gloomy possibility entails that if we want to prevent such a situation from ever coming into being, we must – somewhat ironically – exercise a certain vigilantism of our own. Of course, this doesn’t mean we should stop using social media altogether, only that we should stop ourselves from misusing it, from allowing it to become an instrument of subjugation rather than one of empowerment.