Of Arrest Everyone, Not-so-sutble Tenders, Elephants, Leopards and the Gang that wasn't there.

MisDisMal-Information Edition #7

What is this?This newsletter aims to track information disorder largely from an Indian perspective. It will also look at some global campaigns and research
What this is not?A fact-check newsletter. There are organisations like Altnews, Boomlive etc who already do some great work. It may feature some of their fact-checks periodically

Welcome to Edition #7 of MisDisMal-Information.

Sorry for the delayed edition this week. 

Arrest Everyone!

Once a day, I take a look at Twitter trends to see which hashtags are competing with each other for the trendalympics. I've never been sure what it achieves, but it is a mainstay of the Twitter experience nevertheless. Last week, a very curious phrase was trending on Twitter - ArrestSwaraBhaskar. This seemed to occur in the aftermath of Safoora Zargar's bail petition being denied. Proponents of this trend wanted to Swara Bhaskar to be arrested as well, for a tweet in late January calling for Delhi to 'get on the streets'. But you probably know this already. On Jun 6th, I was able to pull up around 30000 tweets, around the time it was trending at 1 (congratulations?). I also looked at a subset which contained around 35-40 'charged' words. I won't post them here, since this is a family newsletter but I am open to share the spreadsheet with anyone who is interested if you swear an oath not to reveal anyone's personal data.

There was consistency across both the datasets, more than I expected. And there seemed to be no other obvious tell-tale signs of coordinated activity like I saw with male nurses hashtag last week. Other than the obvious one that it was trending at no.1 in a matter of hours.

First, let's look at a spread of users by activity. Note that left axis corresponds to the green line (number of tweets). As you can see, the pattern is pretty consistent.

Full Dataset

'Charged' Subset

Next, the percentage of tweets by when the user accounts were created.

Full Dataset

'Charged' Subset

Again, pretty consistent.

I also looked at associated hashtag clouds for both of them - and it does not make for pretty viewing.

Full Dataset

'Charged' Subset

Btw, if the screenshots are tiny, I've also published these on Tableau Public. Here are the links. Feel free to play around.

 Full Dataset Tag Cloud

'Charged' Dataset Tag Cloud

Oh, and I looked at hashtagify.me for associated tags on 11th June (it works on a relatively small dataset, at least for the free lookups). More people need to be arrested and some still need to be protected, apparently.

There was limited activity on Instagram too, but not much to report. 

Incidentally, some other arrest xyz hashtags were doing the rounds too. For Swati Maliwal and Sonali Phogat. I wasn't able to actually analyse tweets, so I'll leave you with some hashtagify results. While looking into this I stumbled into a network of MRA handles - but that's for another day.



This section is titled Arrest Everyone, so now let's move from the few to everyone. In previous editions, I've spoken about how 'fake news' was being used book cases, lodge FIRs and even arrest people. Well, that trend has continued and shows no signs of slowing down.

I'm compiling these on a Twitter thread, because that's apparently what we do these days. It’s up to 24/25 now.

Notable additions since the last edition:

  • Vinod Dua, who has managed to get himself anticipatory bail. Not everyone will be quite so lucky.

  • Arunachal Police has issued a similar notice to the one Mumbai Police did. WhatsApp admins will be held responsible for "fake news, rumours" on groups. Aside: Can someone explain to me why they thought it would be a good idea to use Comic Sans (or a close relative) for this circular?

  • Udupi's Deputy Commissioner has verbally stated that those spreading 'fake news and misinformation' related to COVID-19 treatment will be booked under the epidemic act. I haven't see a written order though.

  • Hyderabad's Deputy Mayor has filed a complaint with the Cyber Crime Station because 'individuals have been spreading rumours about the health status of Baba Fasiuddin on social media platform'. The News Minute reports that 2 journalists have been booked.

  • Another one from Hyderabad, this time based because someone said the KCR had succumbed to COVID-19.

  • A journalist in Tamil Nadu was booked under the Epidemic act.

  • In Rajasthan, an FIR was filed against 3 staff members of a private hospital. This was the result of statements asking others not to treat muslim patients and that (the poster) would not take X-rays of muslim patients. ‘The three have been booked under Sections 153A (attack upon any religion) and 505 (Statements conducing to public mischief) of the Indian Penal Code and relevant sections of the Disaster Management Act.’ The excuse? I am paraphrasing (see the article for the actual quote) but it sounded like "Oh, it happened while Tablighi Jamaat was in the news."

  • Jammu and Kashmir's Department of Information and Public Relations (DIPR) approved a 'Media Policy-2020' which among other things wants to thwart misinformation, fake news and develop a mechanism that will raise alarm against any attempt to use media to vitiate public peace, sovereignty and integrity of the country.

  • Staying with J&K, an FIR was lodged against unidentified persons for uploading "sensitive posts" on social media which has the potential of harming communal harmony in the Union Territory. It happened after a Twitter account uploaded 'sensitive posts and fake news'.

  • Another one from J&K. There isn't an explicit 'fake news' angle to it, but it is borne out of 'social media criticism'. What was criticised? A verdict by a division bench of the high court. As per the FIR "criticism of the judgment was not only aimed at demeaning the judges of the High Court but also causing disharmony in the general public".

Nothing Subtle about this Tender.

Indian Express reported that BECIL floated an Expression of Interest(EOI)/Tender whose purpose was to "empanel agency(ies) so that their professional services can be utilised for Fact verification and disinformation detection".

Relax, it can't be that bad, can it? CAN IT? Well, Trisha Jalan over at MediaNama points out that this the 8th attempt by the Indian government to directly monitor social media. I guess when life gives you lemons, you make a 'fake news' twist. They were also kind enough to upload a version of the document. Read it and weep. I almost did.

Right on Page 1, BECIL says it has "developed expertise in executing various turnkey based IT projects especially in the areas of Social Media Monitoring, Response management and other related services."


It identifies 5 broad areas of work.

  1. Services for fact verification: This includes 'rapid' fact verification of text, images and videos. Geolocation and verification of visual content. And identify 'Fake News'. 

  2. Services for disinformation content detection and flagging: This has about 18 different items listed under it, but the goal is to analyse 'coordinated disinformation campaigns', analyse automated and bot behaviour (are we assuming all automated activity is bad now?), identifying 'key influencers' behind disinformation and their location. It also covers timeline and sentiment analysis as well as tracking 'disinformation uploaders' across multiple social platforms.

  3. Solutions for rapid verification of claims and images: This calls for 'rapid' querying of facts and images in an existing repository of verified claims using Artificial Intelligence. Let me ignore the 'artificial intelligence' bit for a second. I am bit confused by this one - is the goal to create a database of facts or fact-checks? It reads like it wants to do the former, which sounds...absurd. 

  4. Fake news and disinformation response: Basically the creation of image based fact-checks. Aside - I linked to this in a previous edition,a Shorenstein Centre studyseems to suggest that fact-checks with an audio element do better.

  1. Archival and Storage: Basically - store everything.

What's important to note is that the capabilities covered under 1 and 2, though phrased as a response to disinformation, are sweeping and can be used across many scenarios. And we know what happens when capabilities are developed. Wins are hard to come by, so I'll take a minor one for the fact that the document used 'disinformation' and didn't just smatter 'fake news' everywhere.

The Elephant in the room, and the leopard that doesn't change its spots

It was hard to miss the outrage over the unfortunate case of the pregnant elephant that accidentally consumed fruit laced with explosives. And it was difficult to miss the communalisation of the story. First, the link to Malappuram (which emerged from a reporting error) was debunked.

But leopard didn't change its spots, fictitious muslim names were then assigned blame for the act - which was also debunked.

But we have to pause here and ask - so what if the incident had actually happened in Malappuram instead of Palakkad? Would that make it alright to demonise a whole community?

Speaking of leopards, someone claimed to have seen one in Bangalore. Turns out, it was an old photo.

Yesterday, upon the stair I met a gang that wasn't there.

Anyone remember Antigonish? Absolutely haunting. Anyway, it turns out it may be inspired by Antifa (No, it isn't. That was just a very bad attempt at humour)

Brandy Zadrozny and Ben Collins, who have put out some important stories in the last few days write about locals in Klamath Falls, Oregon declering victory over an Antifa gang that never showed up. And this narrative seems to be spreading.

As the tweet above shows you, Information Disorder can come from the top (ok, vice-top). But the tweet below will show you for the millionth time, that it does come from the top.

On the internet nobody knows you are a ~dog~ ~deepfake~ ~bot~ ~foreigner~ ok, we get it...

New York Times ran a very interesting story asserting that China's Twitter activity was supported by a 'murky' chorus. They stopped short of calling them bots and even managed to interact with some of them.

Vanessa Molter and Renee DiResta from the Standford Internet Observatory published a fascinating essay titled Pandemics & Propaganda: How Chinese State Media Creates and Propagates CCP Coronavirus Narratives. I am including their main observations here, but I do recommend reading it to understand their methodology.

Finding 1: Chinese outlets focus a significant share of their coverage on positive stories and emphasized Chinese Communist Party competence across a range of narratives. 

1.1 Chinese outlets praise the Chinese government response, while U.S. media highlight failures 

1.2 Chinese state media maintained a favorable framing of the CCP response, adjusting the justification for lauding the CCP response over time 

Finding 2: Chinese state media revised, eliminated, and fabricated aspects of narratives to bolster the image of the CCP

2.1 Whistleblower doctor Li Wenliang was first silenced, then celebrated 

2.2 The origin of the coronavirus: First unambiguously Wuhan, then various layers of uncertainty culminating in conspiracy theories 

Finding 3: Chinese state media pays to amplify high-priority narratives 

3.1 Chinese state media ads target English-speakers worldwide 

3.2 Chinese State Media significantly increased ads after January 2020, with most new ads promoting Xi Jinping’s pandemic response 

And if you think we're done with China for this edition, not by a long shot.

Foreign actors and certain third countries, in particular Russia15 and China,16 have engaged in targeted influence operations and disinformation campaigns around COVID-19 in the EU, its neighbourhood and globally, seeking to undermine democratic debate and exacerbate social polarisation, and improve their own image in the COVID-19 context

Apparently they took a potshot at Donald Trump too.

  • China, for its part, has claimed that it was the victim of disinformation.

China, as a victim of disinformation, has always opposed the creation and spread of disinformation by any people or organization, said the Chinese Commission to the EU on Wednesday, in response to the EU's joint communication claiming China has been engaged in targeted influence operations and disinformation campaigns amid the pandemic. 

  • Not the first time they've claimed this, I may have missed including this in the May editions, but China's High Commissioner to Canada had made a similar statement in mid-May.

  • The Manila Times carried a story quoting China's Foreign Ministry spokesperson Hua Chunying as saying:

"Facts have shown that the novel coronavirus is the common enemy of mankind, but what’s more horrible than the virus are disinformation, rumors and slanders spread by people with ulterior motives,"

Around the world

  • The Independent has a llllonng read Tracking the spread of coronavirus disinformation. My favourite bit from the article is about the 'bullshit asymmetry principle' - "The amount of energy needed to refute bullshit is an order of magnitude bigger than to produce it."

  • Marc Owen Jones on Saudi Arabia's disinformation war against Qatar being just the tip of the iceberg.

  • Ray Serrato takes us on a journey of the fake Antifa tweet being shared on Facebook. Sorry, Ray, I had to grab the image.

  • Oxford Internet Institute has been doing a weekly Coronavirus Misinformation Briefing. Some highlights from the latest edition

    • They currently track 142 junk health news websites and 22 state-backed media outlets that are actively publishing misleading information about the coronavirus pandemic—164 in total.

    • All content from all mainstream sources gets the largest amount of total user engagement. But on a per article basis, state-backed content gets over 75 engagements, junk health news gets almost 50 engagements, and average articles from mainstream sources get under 25 engagements.

    • Summed together, 29% of the engagement with non-mainstream information last week was with state-backed content, and 95% of engagement with state-backed content was engagements with Chinese and Russian media content.

    • And apparently, junk health news sources have co-opted the George Floyd Protests to push their own narratives.

  • report by NYU questions the use of the outsourcing model for content moderation that most platforms seem to have adopted, arguing that it leads to marginalisation of an important function.

  • And since we're on content moderation, Engadget reports that Twitter's attempts to flag 5G COVID-19 seem to be over-aggressive, labelling any tweets containing 5G and corona with a 'get the facts' label. Case in point (if the embed does not show the warning, you may have to clickthrough):

  • The UNDP has called for gonvernments to lead the fight against disinformation. Colour me skeptical.

  • Facebook put out its monthly coordinated inauthentic behaviour report covering operations from Tunisia and Iraq.

    • In all, it removed Total number of Facebook accounts removed: 253 Total number of Instagram accounts removed: 240 Total number of Pages removed: 770 Total number of Groups removed: 101

    • Tunisia: It removed 446 Pages, 182 Facebook accounts, 96 Groups, 60 events and 209 Instagram accounts.

    • Iraq: We also removed 324 Pages, 71 accounts, 5 Groups and 31 Instagram accounts.

      Compared to overall activity on Facebook/Instagram, these numbers are.... tiny. But that also goes to show how hard it can be confidently attribute coordinated activity.

And I'll leave you with this thread that looks into 'DM Rooms' to promote tweets/engagement on Twitter. Or what those of us India consider to be 'just another day on WhatsApp'.