Artificial Intelligence has a starring role in India’s 18th General Elections

India: As the world’s largest democratic election in history completes its first phase, AI technologies are becoming increasingly popular with Indian political parties as they campaign for the next few weeks till the results are declared on June 4.

A few days before April 19, a date which marks the beginning of the 18^th general elections in India, the Indian National Congress (INC) released two videos featuring Bollywood stars, Aamir Khan and Ranveer Singh. In both videos, the actors, who each command a large following in India, are critical of the politics and policies of the Bharatiya Janata Party (BJP) to which Prime Minister Narendra Modi belongs; each video states that Modi failed to keep campaign promises, and how he has failed to resolve critical economic issues during his two terms as prime minister. Both videos were Deepfakes and the actors have lodged police complaints against the social media handles that released and shared them, but not before millions of views were generated on Facebook, X (formerly Twitter), Instagram and WhatsApp. This incident is the latest example of how Artificial Intelligence (AI) is making its presence felt in the largest democratic election in history, which will witness approximately 968.8 million citizens, exercising their right to vote to elect 543 representatives to the Lok Sabha (House of the People), the lower house of the country’s bicameral Parliament.

(AI) games parties play

Since February, the two political parties at the heart of Indian politics — the ruling BJP and the INC have adopted AI to play ‘perception’ games against each other; both parties have used AI to generate memes and videos to ridicule their rivals and generate a buzz on social media about their own party. In the previous general election in 2019, “shallow fakes” were effective in manipulating information and perception; these are made manually by using existing simple software tools to alter an image or text, similar to the way daily memes are made. The 2024 elections have witnessed more political parties in India experiment with deepfakes, generating hyper-realistic, photorealistic video or audio files; approximately 40 campaigns have used generative AI so far.

Does AI come with benefits?

The 2024 general elections are staggered over six weeks and seven phases, and is said to be the most expensive elections to be conducted in India — 1.2 trillion rupees or 12 billion pounds. Experts working in the field of political management and communications have opined that the advent of AI will help reduce costs incurred by political parties. For example, AI-generated phone calls have flooded the campaign space, and are expected to reduce expenditure by up to 50 times. The same goes for AI-generated videos. This media can also be tailored according to region and their languages (NB: there are more than 121 mother tongues in India), allowing for a more effective way to communicate with the prospective voters. AI offers political parties to seamlessly transition into digital narratives from the traditional methods.

The government’s approach to (countering) misinformation

Currently, India does not have laws addressing the issue of deepfakes but the Ministry of Electronics and Information Technology (MeitY) recently issued an advisory that asks tech companies to seek MeitY’s permission before launching “unreliable” or “under-tested” generative AI tools. This was met with criticism before the government backtracked and said that start-ups were exempt and that the advisory was only applicable to “significant platforms.” In response, Indian AI companies such as Dubverse, Muonim and Polymath Synthetic Media Solutions, have released an “Ethical AI Coalition Manifesto” — it states that the “integrity of the democratic processes” is upheld by ensuring that “AI technologies are not used to manipulate elections, spread misinformation, or undermine public trust in political institutions.”

Media analysists say that this sort of government advisory seems rushed, ill-researched and anti-democratic, not to mention hypocritical considering that most political parties are using AI; only one or two such as the Communist Party of India (M) has made their use of AI clear in their public outreach posts on social media. While the threat of mis/disinformation looms large over this election season, it has been left to private companies and corporations to take the lead in fighting fake news, deep fakes, and other forms of information manipulation.

Platform responses to electoral integrity

India is reported to have 535.8 million monthly active users on WhatsApp, and in March, Meta launched a dedicated fact-checking helpline on WhatsApp, in partnership with the Misinformation Combat Alliance (MCA). Users will be able to flag deepfakes to a WhatsApp chatbot available in English, Hindi, Tamil, and Telugu, while the MCA will work alongside its “Deepfakes Analysis Unit”, a coalition of independent fact-checkers and research organisations who will identify and verify such content, flag and debunk misinformation. One of the greatest issues with the spread of misinformation in India via WhatsApp is the formation of groups and the way in which false claims or ‘fake news’ is circulated via the forwarding option. Since 2020, Meta has limited users’ ability to forward messages, and last year, the company announced that “any message that has been forwarded once can only be forwarded to one group at a time, rather than the previous limit of five.”

Google has launched Shakti, an India Election Fact-Checking Collective to spot and fight online misinformation. The MCA has collaborated with DataLEADS, who will drive this initiative, along with The Quint, VishvasNews, Boom, Factly, and Newschecker. The network is backed by the Google News Initiative (GNI). The aim of the Shakti project is to assist news publishers with fact-checking information in seven widely-spoken Indian languages — Hindi, Tamil, Telugu, Malayalam, Kannada, Bengali and Marathi — in formats ranging from audio, video and images. Shakti will also train news organisations in the use of Google tools like the Fact Check Explorer, and other advanced verification processes.

What to expect going forward

While these initiatives cast a wide net to trap bad actors who aim to spread disinformation, or clarify false information, there is a not-so-silent question that hovers in the air: are these efforts enough? If not, why?

Digital technologies are proliferating at a speed faster than traditional education in India. In 2022, the degree of adult literacy in the country was about 76.32 percent, mostly men; the global literacy rate for people aged 15 and above is about 86 percent. However, many drop out of school to provide for their families. At the same time, India’s unemployment rate within the ages of 20 and 30 is high, at 44.49 percent. In this socio-economic climate, a vast majority of the population, including the youth, are ill-equipped to comprehend the kinds of technologies that they are being exposed to. There is a significant difference between a tech-savvy population and one that has been trained to think critically, and this divide is impacting the average Indian’s ability to recognise and reject mis/disinformation.