The year 2024 will be remembered as the year when more than 50 countries around the world went to elections, granting the people the right to choose their representatives for a democratic and an egalitarian society. People who have faith in democracy look up to elections as their Cap of Hades to protect just and free society. In a country like India where the internet penetration is high and social media plays a critical role in shaping public opinion, the proliferation of misinformation into the mainstream of our societal fabric, fuelled by lack of digital literacy has made India and its more than 800 million internet users, more vulnerable to the information chaos. Ahead of India's national elections, we witnessed a rising tide of mis-and disinformation often aimed to create chaos to subvert the democratic processes and befuddle the political wisdom of the people.
Little wonder then that the Global Risk Report 2024 from the World Economic Forum highlighted that misinformation is one of the greatest threats that countries around the world are faced with and India, the largest democracy in the world, was identified as one of the nations with the most significant risk associated with disinformation and misinformation. While at DataLEADS, we have been working towards protecting the information ecosystem in India for many years now, given the enormity of the problem, the languages spoken, and the regional diversity, the challenge to protect people from harmful and misleading content during the national elections was enormous.
Disinformation destroys people's faith in authentic news sources.
Shakti Collective came to life with the realisation that there is a need for a concerted and collaborative effort on part of the newsrooms and publishers in the country to put up a formidable fence against misinformation particularly at a time when the spread of misinformation is threatening to undermine the democratic process to an unprecedented degree. The Collective, launched ahead of the 2024 Indian general elections, is by far India's most extensive collaboration to date between fact-checkers and publishers. This initiative, a consortium of over 50 Indian fact checkers and publishers, committed itself to foster accurate reporting of Indian elections by facilitating rigorous fact-checking of election related misinformation. The aim was to create an effective antidote to emerging trends in election-related misinformation. This crucial intervention not only helped fortify newsrooms and publishers towards creating an effective strategy to fight misinformation, but also lent them more capacity through training and resource sharing.
Supported by Google News Initiative, this coalition of Indian fact-checkers and publishers helped enhance the early detection of election-related misinformation and Deepfakes, as well as amplify the dissemination of fact-checks in regional languages.
Importance of Interdisciplinary Teams
The whole initiative was built around an innovative framework, involving multiple stakeholders, 50+ diverse news organizations, including competitors, fact-checkers and technologists who came together to address the common threat of misinformation during the national elections. Every single day during the three month long election period, more than 300 fact-checkers, reporters, editors and technologists collaborated to identify misinformation and combat misleading content, bogus poll surveys, conspiracy theories mostly targeting political leaders, political manifestos, and election processes, including electronic voting machines (EVM) and voter-verified audit trails, among others.
The diversity of the involved teams proved to be a crucial differentiator, and ensured that be it automated alerts system or maintaining the repository and ensuring due compliance, it took every spoke in wheel to come together for the Collective to function like a well-oiled machine.
Power of Collaboration: Amplifying fact-checking networks
India is a diverse country where each region comes with its own challenges with respect to tackling misinformation. This is where the fact-checkers and publishers in the Collective created a significant difference. Newsrooms got easy access to timely fact-checked stories that followed the International Fact-Checking Network's guidelines, and what took fact-checkers a few hours to debunk could be republished instantly. This saved newsrooms crucial time in the busy election season and aided in early detection of election-related misinformation and deep fakes. This amplification across diverse regions and languages helped protect millions of voters from poll-related misinformation and eventually played an important role in saving the integrity of elections, democracy.
More than 6600 fact checks were distributed and amplified during the elections in more than 10 languages, which eventually resulted in the 92% increase in number of fact checks published and 180% increase in regional language fact checks. The remarkable success of the intervention was that 67+ new fact checks desks were activated by publishers in 10 languages across India where editorial teams were actively engaged in countering election misinformation. This election fact-checking effort has been supported by deepfake and synthetic media experts from renowned institutions such as the Indian Institutes of Technology, and other leading AI organizations. These experts emphasized the crucial role that tools and technology play in debunking misinformation, particularly Deepfakes.
Incorporating Language Diversity in Fact Checking:
The Indian election taught us that no one is safe from the misinformation onslaught. While those that speak English have access to fact checks and verified information, the same is not always true of other languages. Regional Indian languages are often under-resourced by fact checking organizations, making it more difficult to tackle misinformation. India is home to 22 official languages and at least 122 major languages, making it especially difficult for fact checkers, to ensure everyone has access to verified information. This diversity acted as a natural barrier against the fighting mis-and disinformation. The spread of false information has had especially dangerous repercussions in the countryside, as misinformation, conspiracy theories and rumours have incited violence and mob attacks, leaving dozens dead in the past few years.
The Collective ensured language amplification by incorporating members from diverse regions. India's leading new agency Press Trust of India (PTI), for example, publishes fact-checks in two languages but when it made its fact-checks available to publishers, they republished the content in 11 more languages. That way the fact checks reached a significant population across the country. Overall we witnessed 180% average amplification of regional language (top 8) fact checks (excluding Hindi and English). Most amplified languages were Kannada with a 302% increase, followed by Marathi (282% ) and Tamil ( 236%).
Language plays a crucial role in shaping our perceptions and behaviours, often in ways we may not fully appreciate. More often the speakers of different languages interpret the world from unique perspectives and that underscores the importance of promoting linguistic diversity in the realm of fact-checking.
Detecting and addressing Deepfakes
Today the synthetic media and Deepfake tools are rewriting the disinformation playbook. There is no better place to study this than India, which, as mentioned already, is one of the nations with the most significant risk associated with disinformation than any other democracy.
In this election year, apart from the usual bad actors like image, audio and video tampering, Deepfake has been seen as a potent enabler of these bad actors in shaping people's perceptions and rupturing the integrity that elections hold. Cloned voices and AI-generated images and videos of political leaders have become a prominent feature, often used to influence voters' decisions and consequently endanger democratic processes. But India's election wasn't infected with deepfake content many feared. The misleading and harmful content was mostly created using relatively cheap and simple techniques such as footage editing or mislabelling to present content in a misleading context. The biggest threat posed by disinformation is that it is not just fake content that makes people believe false things—it's also making them less likely to accept truthful information.
To save the public from deception, the Shakti Collective, established a Synthetic Media and Deepfakes Advisory Council and created daily alerts for newsrooms across India, helping them to be informed about viral misinformation/Deepfakes related to elections. The teams monitored misinformation vectors online and built an online repository of risks, covering 28 Indian states. This eventually served as an early warning detection system and played a key role in identifying and debunking Deepfakes. Based on the survey conducted with Indian fact-checkers who have been at the forefront in fighting misinformation, we conducted a special workshop series in collaboration with WITNESS and India's leading AI scientists on detection approaches and solution of deepfakes including Manual Standard Verification and glitch analysis.
A Blueprint for the future to make collaborations easy
As misinformation continues to batter trust in democracy and its allied institutions worldwide, the lessons learned from the Shakti Collective are more relevant than ever. The initiative's success in India provides a clear blueprint for future collaborations aimed at combating misinformation globally. By streamlining fact-check distribution, pooling resources, fostering interdisciplinary teams, and supporting linguistic diversity, similar projects can be built for a more resilient information ecosystem that upholds the integrity of democratic processes.
Everyone has a role to play in fighting this menace, by verifying and reviewing where they get their information, by ensuring that they don't share falsehoods, and by investing in trustworthy media. The path leading to combating mis-and disinformation is going to require all of us. It is a big and long global fight. One of the most valuable lessons we learned from the Shakti Collective is that when we want to encourage people to do something big, we need to make it easy.
Syed Nazakat is founder and CEO of DataLEADS, a New Delhi-based, award-winning, globally recognized digital media and tech company spearheading a series of fact-checking, digital safety, and information literacy initiatives.