Response to the article published by Rest of World titled "Fake news verification tools fail the test during elections in India
Update: Rest of World has retracted the story titled “Fake news verification tools fail the test during elections in India,” and has replaced it with an editor’s note acknowledging our response and explaining the reason for the retraction. The MCA and its members appreciate the decision taken by Rest of World’s editorial team and their commitment to fair journalistic ethics and professionalism.
An article published by Rest of World titled "Fake news verification tools fail the test during elections in India" on May 24, 2024, raises concerns about the effectiveness of the WhatsApp tiplines run by fact-checkers in India and the MCA’s Deepfakes Analysis Unit (DAU). Not only does the article contain several inaccuracies and misinterpretations of our work, but it also flouts one of the fundamental principles of credible journalism: audi alteram partem (no party should be condemned unheard).
The Deepfakes Analysis Unit of MCA and fact checkers named in the story (also undersigned below) reached out to Rest of World’s team to point out inaccuracies in the reporting and their flawed understanding of how fact-checking tiplines work on WhatsApp. Despite multiple requests sent to the editors to set the record straight, the team at Rest of World remains unresponsive, repeatedly claiming that the responsible individuals have been “in transit” for over four days straight and in others, maintaining radio silence.
The article suffers from these main issues and inaccuracies:
Data Inaccuracies: The data illustrated in the graph, analysing the number of queries sent and responses received, is inaccurate. The numbers recorded by almost all fact-checkers mentioned in the list vary from those in the article. Our tipline dashboards and chat history indicate a significantly higher response rate than what is quoted in the story.
Spammy Behaviour and Vague Queries: Our WhatsApp tiplines follow a clear user flow, guiding users through each step to ensure claims are submitted correctly and reach our dashboards. However, Rest of World’s reporters did not adhere to this process and inundated the tiplines with media items in a manner flagged as spammy by the system. As a result, many of these queries were filtered out. Their reporters often failed to specify their inquiries or provide context despite our requests for clarification. For instance, we received multiple identical queries asking, "Can you verify this?" within a short timeframe. In one instance, 11 queries were sent within 7 minutes without following the established flow. Additionally, lengthy videos were submitted without specifying what aspect needed verification.
Condemned without being heard: Rest of World failed to contact all 11 named entities for comments. In a rush to publish, they provided some fact-checking organisations with less than 24 hours to respond to exceedingly vague requests. During this crucial election period, our primary obligation is to assist authentic tipline users and debunk viral and impactful misinformation online.
Misrepresentation of what tiplines do: We believe that the core issue with this article lies in its foundation on a false premise: that tiplines function as "AI detector chatbots," a notion they emphasise in their article as seen in the image below. Our tiplines work across the entire spectrum of dis/misinformation and are not specifically aimed at AI detection. Only the DAU’s tipline focuses on AI generated audio and video. Even then tiplines don’t do real-time and immediate fact-checking or detection. This significant misunderstanding coupled with erroneous data has resulted in sensational yet flawed conclusions.
Misattribution: Some of the quotes have been misattributed and persons misquoted, and despite clear communication pointing out inaccuracies, from multiple entities named in the report, no changes have been made so far. One error was made twice, as the MCA had corrected Rest of World previously on another story of theirs.
Lack of basic research: The story also confuses timelines of the launch of several initiatives relating to WhatsApp tiplines, and assumes all fact-checkers work on the same tech stack, without doing basic research to understand who uses what.
The MCA’s Deepfakes Analysis Unit and the fact-checking organisations named below reached out to Rest of World in good faith to address the inaccuracies and call for corrections. The (lack of) response has been disappointing, aimed at buying time, so the flawed story can run its course and damage our respective reputations. We urge them to take urgent action, investigate what went wrong, and be transparent to all their readers on how they went wrong.
This statement is made on behalf of:
Bharat Gupta, President, Misinformation Combat Alliance
Rajneil Rajnath Kamath, Vice President, Misinformation Combat Alliance
Pamposh Raina, Head, Deepfakes Analysis Unit
Jatin Gandhi, Executive Editor, Vishvas News (Jagran New Media)
Rakesh Dubbudu, CEO, Factly
Jency Jacob, Managing Editor, BOOM
Dheeshma Puzhakkal, Fact Check Lead, NewsMeter
Jaskirat Singh Bawa, Global Head of Editorial Operations, Logically Facts
Ruby Dhingra, Managing Editor, Newschecker
Ritu Kapur, Managing Director, Quint Digital Limited
Rahul Namboori, Editorial Head, FactCrescendo
Bal Krishna, Executive Editor - Fact Check, India Today
Saurabh Shukla, Editor-in-Chief, Newsmobile