Tow Center

Terrorism bred online requires anticipatory, not reactionary coverage

March 20, 2019

Journalism has a panopticon problem. Just like the watchman in the center of Jeremy Bentham’s theoretical circular prison, we can now surveil a vast amount of human activity on the internet. Either through watching users reveal themselves on social media, or by digging into the carelessly secured digital footprints we all leave behind, the amount of information journalists and others can now “see” outweighs what we could and should report. Unlike Bentham’s penitentiary, in which the knowledge they are being surveilled limits the brazenness of the prisoners, the easy publicness of actions on social media has not served to discourage bad actors. Instead, it has empowered them to shock and radicalize more widely than would have otherwise been the case.

In the wake of the pre-meditated murder of 50 Muslim worshipers in mosques in Christchurch, New Zealand last week, journalists have found themselves in a familiar place; we are back into the loop of self-examination in terms of how to report about acts of terrorism and on those who commit them. In the Christchurch killings, the gunman conducted mass murder within the framework of a social media campaign. A long screed was posted online on an easily accessible message board, explaining his motivations and namechecking a long list of inspirations—some obviously ironic, some apparently sincere. Two Twitter accounts were set up to help spread news of the atrocity, and he used Facebook Live to stream video of his actions in real time, during which he dropped references to aspects of online political culture across all parts of his deadly performance.

Individuals, trained by the platforms to understand how to influence and publish, are now well versed in creating material the platforms find difficult to purge.

Facebook, Twitter, and message groups struggled to contain the material with limited and late success. Both Facebook and YouTube talked about the difficulty of removing the millions of repeated uploads from the atrocity. Neal Mohan, YouTube’s chief product officer, made the problem sound unmanageable, as a version of the video of the killing was uploaded as often as once per second. “This was a tragedy that was almost designed for the purpose of going viral,” Mohan told The Washington Post. In a statement, Facebook said there were 1.5 million attempts to upload the video, 1.2 million of which were prevented “at upload.” These are impressive numbers, but the algorithms weren’t enough to keep the virality at bay. We saw in the response of the platforms the limits of artificial intelligence to quell the path of information— troubling, given that Mark Zuckerberg has in the past stated that he sees AI as a major path forward for content moderation.

Individuals, trained by the platforms to understand how to influence and publish, are now well versed in creating material the platforms find difficult to purge. Information, it seems, will get out. This raises a question for the press about how journalists ought to treat newsworthy but highly sensitive events. While the ambition is always to have the best possible reporting in difficult situations, there is inevitably debate about where the line should be drawn in terms of coverage and focus—now, that balance must also take into account the fact that the press no longer controls what information is available.

There is a whole literature now on what best practices look like for journalists and editors reporting as live events of this type. Many point to academic studies such as this, from 2015, linking a “contagion” effect to mass killings, and to thoughtful guidelines on how to cover mass shootings responsibly. The Tow Center, which I direct, produced a series of reports in 2016 about the interdependence of terrorism and the media and political response. Responsible reporters ought to have the basics imprinted on their subconscious: Do not report facts until they are verified, do not focus on the perpetrator over the victims, do not use sensational language that might glamorize the terrorist.

In response, Jacinda Ardern, the Prime Minister of New Zealand, told the New Zealand parliament she would not repeat the gunman’s name. She may have been taking a page from US coverage of mass shootings: In the wake of the shooting in Aurora, Colorado in 2012, two parents set up a campaign challenging the media to stick to a policy of ‘no name, no photo, no notoriety.’ Similarly, last year, Joan Donovan, an online speech researcher, and danah boyd, director of Data and Society, presented the case for journalists to practice “strategic silence” when considering issues of amplification of online speech and ideas.

Sign up for CJR's daily email

READ: A mosque massacre is livestreamed

Since journalism is a citational field (journalists need to show evidence that something happened), Donovan says, the temptation is to link to sources or post material, such as the Christchurch shooter’s document, whole without context. But this time, there were clear signs after the Christchurch shooting that professional journalists were thinking more carefully about the focus of coverage: “I haven’t seen, for instance, journalists turn to white supremacists—who they now have in their contacts—to speak about this, and that is to be commended.” And linking to far-right sites and the publishing of raw material was less obviously a feature of mainstream coverage, she adds.

Strategic silence is a difficult concept for journalists, and it has its limits. In the immediate aftermath of Christchurch, I searched government-friendly media outlets in countries with authoritarian and Islamophobic leaders, such as Russia, Hungary, and the Philippines. While all carried coverage of the attack, often it was not the lead item—and in most cases the details of the attacker and his motives were left out or vague. Who does this serve? Right-wing talk show host Ben Shapiro urged his followers to ignore the identity and ideals of the gunman, so as not to spread the hateful message. Others noted that it serves the interests of the far-right commentariat who have regularly posted Islamophobic commentary to deflect attention away from the uncomfortable proximity of the deranged meanderings of white supremacists to their own sentiments.

The issue of how much reporting attention we give to the micro-details of attacks which use social media or which are widely captured by witnesses with cell phones is another difficult line to navigate.

The algorithms weren’t enough to keep the virality at bay. We saw in the response of the platforms the limits of artificial intelligence to quell the path of information.

There are others who think reporting blackouts or over-zealous self-censorship are not helpful in combating terrorism or violent extremism. Bellingcat, a website which specializes in open source intelligence (OSINT) reporting, produced a comprehensive and contextualized report on the motives and movements of the Christchurch killer before he carried out the attacks. The piece contextualized his use of internet image boards and the references he made in his writings and video.

“His manifesto is not dangerous to the average reader,” Bellingcat’s Robert Evans, who wrote the piece, tells me. “The people who will be most influenced by the details of this are people who are already in the dark corners of the internet and already know.” Evans would not have posted Tarrant’s writing whole or uncontextualized, “but pointing out things he feels are important, what radicalized and influenced him is important.”

Evans feels that a lack of coverage of far-right terrorists has made it easier for platforms to ignore those terrorists’ content, and contrasts that willful ignorance with the near erasure of ISIS and radical Islamist activity from the open web by the very same platforms, at the urging of activists and policymakers. “ISIS and their tactics are very much the blueprint for white nationalists. What has effectively minimized the online presence of ISIS though is the political will to do so.” (In a Tow Center report from 2016, Rafia Zakaria writes that US law has no way of categorizing domestic terrorism as terrorism, legally.)

READ: New Zealand massacre: Journalists divided on how to cover hate

In the panopticon, where information is readily available to journalists and nonjournalists alike, there are other concerns that journalists should be routinely aware of. The new term is stochastic terrorism, which describes an unanticipated act of terror committed by a seemingly random individual. Where there is a saturation of inflammatory rhetoric about ideology or particular groups, the theory goes, it becomes statistically likely that some lone wolf will take the bait. The terrible events in Christchurch bear the hallmarks of stochastic terrorism prompted by far-right influence and endless Islamophobic themes repeated by prominent politicians and elements of the media. For the Christchurch killer, as for Norwegian far-right terrorist Anders Breivik—whom the Christchurch killer cites by name—the route to terrorism was close to the mainstream, and reflects a constant type of Islamophobia present on news websites with anti-Muslim agendas such as Breitbart and the Daily Mail.

Evans feels that if the white nationalist industry were less profitable to social platforms, and if image boards such as 8Chan were taken more seriously as a root of terroristic threat, then the many clues and postings made ahead of the attack would have been caught earlier. The flip side of strategic silence is that the threat of a different kind of terrorism is minimized, and that the platforms, who have never acted to pre-empt anything that has not previously attracted coverage, continue to treat extreme and violent views as a tolerable type of speech.

If there is anything to be learned from the incredibly dark events in New Zealand it is that the platforms as well as the press are now being held accountable for their reactions and policies in terms of how material enters the public square. “One thing we have to assume is that platforms are editing and shaping what we see. The ship has sailed in terms of thinking we ever see unfiltered speech, even search is curation,” Donovan says, “It might be that this is the incident that responsibilitizes the platforms.”

Journalists and technologists in the meantime should be using every incident which fuses the use of both mainstream and social media with devastating outcomes for ordinary citizens as a chance to reflect on their own role and responsibility in a changing environment.

ICYMI: WSJ reporter explains why he was fired

Emily Bell is a frequent CJR contributor and the director of Columbia’s Tow Center for Digital Journalism. Previously, she oversaw digital publishing at The Guardian.