By: Samir Krishnamurti
We are living in an age of overload. We have too much information coming at us from all quarters, and today’s social-media driven fast news cycles certainly doesn’t help matters much, let alone allow us to get a full picture of why things are happening the way they are happening.
With platforms like Facebook and Twitter becoming a major conduit for news distribution and consumption, so have they been at the receiving end of criticism for enabling the broadcast and amplification of viral propaganda and misinformation, as well as disrupting electoral processes on a scale that was previously unimaginable.
“While many fields have been disrupted by automation and computation, few have converged as abruptly and as publicly as software engineering and journalism,” wrote Emily Bell, Director of Tow Center for Digital Journalism at the Columbia Journalism School, adding, “The news media is witnessing its business models and production processes being remade by Web publishers and search engines. Every major news event in the world, from bombs raining down on Aleppo to the late night tweeting of presidential candidates, is broken through social media and seen through our luminous mobile phone screens.”
But the tech-journalism divide isn’t just crumbling with regards to ushering newsrooms into the digital era. It is also about leveraging technology, and by extension, social media, to use them as an effective tool for gathering factual assertions and evaluating their accuracy. Of course, when such statements are put forth by politicians on a public platform, it is only a matter of time before they are taken apart and analysed for any untruths.
For example, both The Washington Post and The New York Times have taken a close look at every single ‘suspect statement’ uttered by President Donald J. Trump, putting together an exhaustive database of lies. This trend of fact-checking was sparked in the United States by initiatives like PolitiFact and Snopes, and has since caught on in other parts of the world.
Duke University’s Reporters’ Lab, which maintains a global map of fact-checking organisations, puts the current active number of such operations at 162. Irrespective of how and where they operate, fact-checkers subject all statements made by politicians and public figures to rigorous analysis and research, before rating them as true, false, or misleading.
While fact-checking undoubtedly has an important role to play in debunking claims and deter the dissemination of misinformation, being transparent about their methodology goes a long way towards establishing credibility. No method, whether computer-assisted or fully human-driven, is completely devoid of any shortcomings, and mistakes do happen. PolitiFact, for instance, was forced to retract a story in light of new evidence last month, highlighting the dangers of rating claims without full facts in hand.
Yet fact-checking remains a crucial tool for improving accountability. A research published in 2013 under the title “The Effects of Fact-Checking Threat” found that “legislators who were sent reminders that they are vulnerable to fact-checking were less likely to receive a negative PolitiFact rating or have the accuracy of their statements questioned publicly than legislators who were not sent reminders.”
But compounding the difficulty in assessing claims is the virality of social media. As misinformation and disinformation continue to spread like wildfire on platforms like Facebook, Twitter, and WhatsApp, it has been as much a challenge for the companies as it is for fact-checkers to contain them before it’s too late. Facebook has partnered with many third-party fact-checking organisations, like Boom in India, to help tackle the problem. In addition to reducing the distribution of a story rated as false by the fact-checker, the social network has taken it upon itself to monitor pages and domains that repeatedly share such debunked news stories.
This strategy has not been without its fair share of criticism too. Earlier this April, Facebook was criticised by the Philippines government for choosing fact-checking platforms that they say were biased against President Rodrigo Duterte. The company’s partnership with Snopes backfired in the United States as well, after users who were sympathetic to false narratives continued to share the story despite it having being proven false.
While Facebook continues to fine-tune its approach to fake news on the social network, WhatsApp still continues to be a far more complicated Gordian Knot to untangle. After more than a dozen people were killed in a spate of mob lynchings fuelled by WhatsApp messages in India, the company took the drastic step of limiting forwards to just five chats.
Fact-checkers, for their part, have gone on to establish their own dedicated channels on WhatsApp for people to send rumours and uncorroborated claims, which are then investigated and rated. One such initiative has been First Draft’s Compova, which brought together “journalists from 24 different Brazilian media companies to identify and explain rumours, fabricated content, and manipulation tactics that might influence the 2018 presidential election campaign.”
U.K.-based Full Fact, in contrast, takes an automated approach to fact-checking, parsing claims from British media and politicians, and then cross-referencing them with data from the Government, research institutions, and independent experts to label them as inaccurate or correct. Similarly, a research project out of Duke Reporters’ Lab combines journalism and computer science to work on an app that provides pop-up fact-checking in real-time by matching “spoken claims with fact-check articles that have already been published” in places like The Washington Post, PolitiFact, and FactCheck.org.
Yet other hurdles do remain. Election campaigns have been found to launch their own partisan fact-checkers while adopting the style and methodologies of independent fact-checkers, thus weaponizing the medium to gain political mileage. Another important question is the one of bias. Who is checking the fact-checkers to see if there is a partisan bias in their wording? Along with it comes the danger of selective facts, which are ‘true’ facts that tell only a part of the story, and therefore can lead fact-checkers to draw a different conclusion, one that may be totally erroneous if all key pieces of information were present.
Fact-checking therefore not only entails meticulous fact gathering, but also in ensuring that the corrections reach the intended audience so that fake news can be effectively suppressed. While research indicates that stories rated ‘false’ tend to spread faster and wider than those rated ‘true’, it has also been found that Twitter users who are tweeting misleading claims are more likely to accept corrections if they come from friends and individuals who follow them, but are less likely to be accepting of them otherwise.
All of this points to a new landscape where fact-checking and detecting misinformation in digital media becomes a collaborative, crowdsourced effort involving journalists, technologists, civil society, and the general public to bridge conflicting echo chambers so that opposing opinions can be shared in a meaningful manner. Algorithms can be designed to detect different forms of false information, including verifying fact-checking websites themselves to see if they are deceptive and aping another reputed fact-checking organization, while leveraging crowdsourcing as an option to detect and mitigate the spread of misinformation.
Whatever be the approach to countering fake news, the most important takeaway is to be factually correct. “We should hold on publishing until we are sure that we have all the facts ourselves. That our conclusions are solid,” said Bill Adair, founder of PolitiFact who is now the Knight Professor of Journalism and Public Policy at Duke University, where he specialises in journalism and new media. “It can be tempting when you are working on the Web to find one website that confirms your suspicion that the statement is false or true and quickly publish. But you need to be very thorough and confirm all the possibilities.”