Alt News co-founder Pratik Sinha on the fake-news ecosystem in India

Courtesy Pratik Sinha
08 March, 2019

Pratik Sinha is a software engineer by training, with a specialisation in wireless and embedded systems. Two years ago, he co-founded Alt News, one of India’s only fact-checking portals at the time. He has since become a key figure in the social-media age of news.

In February this year, Tushar Dhara, a reporting fellow at The Caravan, interviewed Sinha about the crisis of misinformation and the kind of fake news that goes viral in India. Sinha also spoke about the process of fact checking at Alt News and the steps that Facebook is taking to combat the spread of misinformation. In the context of the upcoming Lok Sabha elections, he observed, “We cannot stop bad actors from pushing out misinformation, you can only control the velocity.”

Tushar Dhara: Can you describe the evolution of the fake-news ecosystem in India?
Pratik Sinha: The right wing organised themselves on social media much before any political party [with a different ideological leaning]. The only other political party that organised themselves on social media prior to the 2014 general election was the AAP [Aam Aadmi Party]. When you organise yourself on social media, you have a network of people you can reach out to.

Back in 2013–14, there were multiple factoids and pictures on social media saying, “This is Gujarat,” “That is Gujarat,” because the Gujarat model was the focal point. For instance, what was claimed to be the Ahmedabad BRT [Bus Rapid Transit system] was a picture from China. Back then, because the right wing was better organised, there was more disinformation coming from that part of the ideological spectrum.

Fast forward to 2016—the launch of [Reliance] Jio was announced and various telecom companies started dropping their rates, making internet data more affordable. At the same time, there was a huge influx of Chinese phones—both smartphones and feature phones—costing Rs 2,000 that come bundled with WhatsApp and Facebook. Owning a phone—including the cost of the device and data—became cheaper. The evolution of the first few fake-news sites, like, Dainik Bharat and Postcard News, can be mapped to the increased availability of data and cheap phones.

Recently, when Abhishek Mishra—he specialises in anti-BJP fake news, and is based out of Madhya Pradesh—was arrested [for publishing posts that hurt religious sentiments], I was told he has a large office out of which he operates, he employs people and puts out misinformation. These are young people who realise the potential for misinformation from a financial point of view. Mishra used to do pro-Modi videos. Then he switched and went to the anti-BJP side because he thought there was more scope on the other side.

[There is an] increase in people consuming data [and] a huge section of the population has very little internet literacy. Political parties realised this and made putting out misinformation an organised thing. In the case of the arrest of Mahesh Hegde from Postcard News, senior party functionaries [of the Bharatiya Janata Party], including union minister Anant Hegde, defended him.

TD: In India, do you see misinformation being spread by parties other than the BJP?
PS: Absolutely. We [Alt News] started in February 2017. The first year was completely dominated by the right wing in terms of misinformation put out. But now, there are several pages that do pro-Congress propaganda. If AAP puts out misinformation, they are usually apologetic.

Recently, we put out two stories debunking misinformation against the BJP. One claimed that Yogi Adityanath [the chief minister of Uttar Pradesh] was laughing at those who died at Pulwama during a condolence meeting. [On 14 February 2019, a suicide bomber drove a car filled with explosives into a security convoy of Central Reserve Police Force personnel in Jammu and Kashmir’s Pulwama district, killing at least 40 people.] It was actually a video of him [attending] ND Tiwari’s funeral [a former chief minister of Uttar Pradesh and Uttarakhand].

TD: You recently edited a Google document that contained tweets that supporters and members of the BJP then posted on their accounts. How did you gain access to the document?
PS: I monitor multiple WhatsApp groups that I have been able to infiltrate. [On one such group,] I found this document and the global edit option for it was open. I edited it and they started tweeting it out. I was making videos [of this] because I wanted to put it out. The thread went viral and the right wing was absolutely stunned.

TD: How does fact checking at Alt News work?
PS: At Alt News, we first monitor misinformation. Disinformation is news that you know is false and still circulate, while misinformation is where the sender is not aware if something is true or false but decides to forward it anyway. So, disinformation is a deliberate attempt to spread false news.

In the case of Facebook, we use a tool called CrowdTangle [a tool that tracks how content spreads across the internet]. We use it for monitoring different Facebook pages—at this time we monitor 600–800 pages, which, at some point or the other, have put out misinformation and are on either sides of the ideological spectrum. Using CrowdTangle, we are able to see what kinds of conversations are happening and which ones may go viral. We also monitor Twitter using TweetDeck [a Twitter management tool]. We have a list of people who tend to tweet misinformation frequently.

When we identify the posts, we start looking [to identify] whether there is an element of misinformation or not. When we determine that there is misinformation, and that it is viral enough, we step in and do a fact-check. The virality is important—if there is misinformation that is not being circulated widely, we don’t want to write about it because it might have a reverse effect and people may start circulating it.

Many people tag us on Twitter and Facebook and send us stuff on WhatsApp—people are alerting us, and that has increased over time.

TD: What kind of fake news goes viral in India?
PS: In case of the Indian social-media ecosystem, a large majority of misinformation is in the form of misleading images and videos—a video or image with a text blurb. That is 70 percent of misinformation, especially the communal kind of misinformation.

Misinformation in the Indian context is mostly of two types—political and medical misinformation. It is very difficult to monitor medical misinformation because it does not have a pattern, it is not motivated misinformation. For instance, the claim that papaya leaves or seeds help fight dengue by increasing the platelet count—when something like this is shared, it doesn’t follow a definitive pattern.

In case of political misinformation, there are two–three motives. One is an organised attempt to target minorities and show them in poor light—the attempt is to show that they are responsible for many things that are wrong in the country. Another motive is to target individuals—when the Gurmehar Kaur issue became big, [Kaur led a social-media campaign in February 2017 called #StudentsAgainstABVP, following which she received several rape and death threats on social media] she was targeted using a false video where a girl was shown dancing and drinking, which is looked down upon by many sections of Indian society.

When the Padmaavat protests were happening and the Karni Sena was out on the roads [because they claimed that the Hindu princess Padmavati’s story was being distorted in the movie], a school bus was attacked in Gurgaon [a city in Haryana]. After the arrests of [18] people, a message containing five random Muslim names went viral, [which stated that they had attacked the school bus.] There’s a pattern—the order of these five names will never change during the duration of the misinformation.

Another example is the Narendra Modi UNESCO award forward, where the misinformation got repeated over thousands of accounts. [In June 2016, a message claiming that UNESCO had declared Modi the best prime minister of the world went viral.] This kind of misinformation does not hurt anybody. But the [Karni Sena kind of] misinformation intends to create or deepen biases in society.

TD: The 2019 general elections are coming up. How can we safeguard the political process from fake news?
PS: We cannot stop bad actors from pushing out misinformation, you can only control the velocity—that is where social-media platforms come into the picture. In case of Twitter, there is no way to report misinformation. Now, [on] Facebook there’s a way to report misinformation as they have been under enormous pressure post the Cambridge Analytica [incident, which indicated that the political-consulting firm had accessed the data of millions of Facebook users].

Facebook started contacting various fact checkers, and now they have fact checkers on board [in India]. They have created a system—I have not seen this firsthand, I have only heard about it. It is like a console and you say these are the URLs of the misleading posts, this is the corresponding fact check, and you write a brief summary.

TD: In February this year, Facebook announced that it was adding five new partners to its third-party fact-checking network in the country. What do you think of these partnerships?
PS: The latest fact checkers include India Today, which started a fact-check desk. Vishwas News is the [Dainik] Jagran’s fact-checking initiative. An institute, [the International Fact-Checking Network, a unit of Poynter Institute, which is a non-profit school for journalism] is giving them certificates after looking at whether they have any direct political links, or ideological leanings. Facebook has made it a criteria—if you have a certificate, we can have you on-board. I do have certain concerns about some of those who have been brought on-board, but they have got the certification. From Alt News’s point of view, we are going to observe these fact-checking outlets and see what they do. It’s premature to comment on them because we haven’t studied them in detail.

TD: Which of the fact checkers do you have concerns about and why?
PS: I can’t comment on their work because I haven’t looked at it. But for example, NewsMobile’s founder is a partner of Republic TV, and you know what Republic TV has done. It immediately raises some red flags. But I don’t have any data to say NewsMobile has been biased. I don’t even think many people know NewsMobile as a news organisation, or if any of their stories have been widely talked about. Vishwas News is another partner [of Dainik Jagran] and Dainik Jagran has put out stories [such as one that said] there was no rape in Kathua. That is a concern.

TD: Do you think forming such partnerships is enough to tackle fake news in the country?
PS: I don’t think it’s enough. I would like to add that it’s a good start to create fact-checking partnerships which are ratified by the International Fact-Checking Network so that it is not Facebook which is deciding what is the truth and what is not.

My criticism is that it is a wholly or largely a manual process. Facebook could have differentiated itself—it should have introduced tech into the fact-checking process. I am [not] talking about WhatsApp—it will have to undergo many changes to counter this.

When the Amritsar train tragedy happened during Dussehra—a train ran over several devotees—a message went viral after an hour, saying the name of the train’s driver is Imtiaz Ali. There were three–four versions of this message doing the rounds with the claim that because it is a Hindu festival, Imtiaz Ali ran the train over devotees. Now, when you fact check this, you have to list the offending URLs. When something goes viral, a human fact-checker can only input the major offenders—it is humanly impossible to flag thousands of links. Incidentally, one of Facebook’s oldest fact-checking partners recently ended their partnership and cited this exact reason—it was a very manual process.

For example, [Facebook] has algorithms which recognise copyright infringement in videos and images. If a user puts out content, it will compare it against a database to figure out if your video matches any videos in there—if it does, it is deemed a copyright infringement. The technology that Facebook uses to figure out copyright infringement is the same technology that it needs to use [to fight fake news].

TD: What can Indian newsrooms do to call out fake news and create a fact-checking ecosystem?
PS: The objective of fact checking is to create a doubt about what they may be forwarding because a lot of people forward things without giving it a second thought. News channels or FM [radio] channels could have a small segment everyday where they are debunking stories—if people [follow] it regularly, they will start questioning fake news. We have seen it on a smaller level where people on social media are saying, “Is this true?” Education could be another thing—I think [Kannur, a city in Kerala] is already putting it in its curriculum for people to identify misinformation. For me, media starting to fact check is a good thing.

TD: Do you think this surge in fake news in India is a reflection of how credible journalism fails to reach the masses?
PS: There’s obviously an issue with channels such as Republic TV, Times Now, Zee News or Sudarshan News. The narrative they put out is problematic—it feeds into not just misinformation, but a lot of things. For example, right now Kashmiris are being targeted—we need empathy that they are being targeted. Journalism is deepening political and religious biases, and when you have such strong biases, you are likely to fall for misinformation.

This interview has been edited and condensed.

Correction: An earlier version of this article incorrectly stated that a picture from Singapore was floated on the internet with the claim that it is from Ahmedabad. The picture was in fact from China. The Caravan regrets the error.