13% Images on Public-Political WhatsApp Groups Misinform: Study

The three main categories of image misinformation on WhatsApp includes: Out of context, memes and photoshopped.

(Photo: Shruti Mathur/The Quint)

Video Editor: Deepthi Ramdas

Nearly 13 percent of the images that are circulated on public-political WhatsApp groups in India peddle misinformation. This was found in a study titled ‘Images and misinformation in political groups: evidence from WhatsApp in India’ conducted by two MIT researchers between the end of 2018 and mid-2019 – a period that included events like the Pulwama attack, the Balakot airstrike and the Lok Sabha elections.

“One of the main findings is that roughly one in eight images that are shared on these public groups, they are misinformation. They contain some kind of factual inaccuracy. The second thing is that we are trying to find and understand how these images are being misused.”

Kiran Garimella, Reseacher, MIT

According to the study, there are three main categories of image misinformation:

1. Out of context: This essentially means an old image which has been taken out of context and re-shared.
2. Photoshopped: These are simple doctored images.
3. Memes: These are mostly fake quotes attributed to prominent people or made-up statistics.

These three categories of images constitute over 70 percent of misinformation in the sample.

(Source: Deepthi Ramdas)

What Can be Done to Tackle Misinformation?

Unverified images and videos circulated on WhatsApp have led to assault and even lynchings in India. In fact, in 2018, WhatsApp said that India tops the world in forwarding messages.

However, another study published in June 2020 showed that ‘nudging people to think about accuracy’ could actually help people in deciding what they want to share.

Garimella said that they have showed in their study that having something on your device to fact-check an image before you forward it, can help curb the spread of misinformation.

“The simple solution that they can do is to have something on the device where before you post something you can already fact-check it. And this is technically possible without breaking encryption, that's what we show in our work, which actually helps in preventing a lot of misinformation from being shared,” he added.

The study also mentions that the machine learning models are better at identifying the three types of misinformation which includes: out-of-context images, photoshopped images and memes.

How Have Images Been Misused in the Past?

As mentioned in the study, over the years, images have been used to misinform and mislead people.

The Quint’s WebQoof team has debunked various such images in the past where either old pictures were used out of context to spread misinformation, memes were passed off as news or photoshopped images were circulated.

This image was found to be old and had nothing to do with the Indo-China standoff. (Source: Facebook/Screenshot)

Recently, during the Indo-China standoff, multiple images were circulated on the internet with claims about casualties on the Chinese side and higher number of casualties on the Indian side. However, most of them turned out to be old and unrelated.

Similar trends have been noticed earlier during the anti-CAA protests, Ayodhya verdict, abrogation of Article 370 and other major events.

(Photo: WhatsApp/Altered by The Quint)

(Not convinced of a post or information you came across online and want it verified? Send us the details on WhatsApp at 9643651818, or e-mail it to us at [e-mail] and we'll fact-check it for you. You can also read all our fact-checked stories here.)