If India’s most recent elections proved anything, it’s that the “world’s largest democracy” may yet still be worthy of that name. The results of the six-week cycle, which saw more than 640 million voters turn out across the subcontinent, dealt a steep blow to the demagogic Prime Minister Narendra Modi and the ruling Bharatiya Janata Party, which lost its single-party hold over the majority of seats in India’s lower house of Parliament, the Lok Sabha. The Hindu nationalist BJP still snapped up most of the electorate’s votes, and Modi is bound to continue as head of state, but he and his cronies will no longer enjoy the untrammeled federal power they have held for the past half-decade (forget the 400 seats they’d aimed to nab in this election).
The INDIA Alliance, a multiparty coalition of opposition candidates, has flipped dozens of seats, and the BJP will have to rely on parliamentary allies—who already desire some significant changes to Modi’s Hindu supremacist governing style—for a majority. In light of the clean sweep Modi and Co. achieved in 2019, the increasingly authoritarian grip they brought to India’s institutions, the economic depression that has afflicted Indians in the Modi years, the protest movements that resultantly flared up, and the formerly BJP-supporting constituencies the party lost this round, it’s hard to view this outcome as anything but a popular rebuke of Modi’s antidemocratic excesses.
Those excesses include the powerful tech apparatus the BJP has employed since Modi first campaigned for national office in 2013. Popularly recognized as “India’s first social media prime minister,” known for cozying up to Silicon Valley magnates, and boosted mightily by wealthy tech-world allies, Narendra Modi promised to be the avatar of “Digital India,” with a presence as ubiquitous online as off. During this time, more of India’s 1.4 billion people were gaining access to cheap phones, mobile data, and internet bandwidth.
That virtual face had a hateful glow starting early on, as the BJP hired volunteer stooges to target Modi critics, journalists, members of non-BJP parties, Muslims, and other minorities with harassment campaigns, inflammatory disinformation, and death threats. These proudly self-identified trolls deluged Facebook, Twitter (pre-X), and especially WhatsApp with pro-Modi propaganda, and eventually built out a sophisticated army with strategic cells and on-the-ground foot soldiers deployed across the country. They benefited, too, from either knowing inaction or blatant acquiescence from American platforms that wished to keep doing business with India’s massive consumer bases—even as these BJP initiatives sparked grisly real-life pogroms.
In a more recent example: When Elon Musk took over Twitter, the CEO was happy to oblige Modi’s aggressive requests to ban or obscure accounts of dissidents and even countries perceived as unfavorable to Modi. Musk also restored the accounts of nationalist officials and activists who’d been previously booted for violating Twitter policy. (YouTube and Facebook haven’t done much better, although Meta has stood firm against the Indian government’s request to break WhatsApp’s encryption.)
These social media horrors, combined with the government’s censorship efforts, fostered an informational environment in India ripe for exploitation by artificial intelligence. The country’s right-leaning platforms embraced the A.I. moment in early 2023: The hate-filled Twitter clone Koo incorporated ChatGPT into its text boxes, and an engineer based in Indian tech hub Bangalore crafted a “GitaGPT” app that spawned a proliferation of “talking” Hindu-deity chatbots praising Modi and promoting violent acts on behalf of Hinduism. When women wrestlers rose up later that year to protest a coach (and BJP politician) who they claimed had sexually harassed many of them, party acolytes used A.I. to edit an image the wrestlers had posted to social media while they were held in police custody—making it appear as though they were smiling instead of showing their disgust with the system.
During the COVID-19 pandemic, a former student activist began a popular, nonideological contract service called the Indian Deepfaker, which was soon “approached by a political party to translate the video of a politician’s speech in Hindi to a regional language,” according to the Indian fact-checking site Boom. After the Indian Deepfaker went viral for incorporating Kangana Ranaut—an actress and far-right provocateur who just won a Lok Sabha seat for the BJP—into a cocked-up Barbie trailer, other A.I. enthusiasts began deepfaking Indian politicians, even creating voice clones that could be translated into some of India’s most common regional tongues. The newfound capability presented an advantage for national politicians versed primarily in India’s official language, Hindi, to transmit communications to voters in southern states in other languages and dialects, like Kannada and Malayalam.
Soon enough, a guitar-playing A.I. Modi was singing in Telugu on Instagram. The right-wing Insta profile Hokage Modi Sama paid to boost hundreds of Midjourney-generated deepfakes that made Modi resemble figures from Hindu mythology. In Tamil Nadu, the chief minister used an A.I. re-creation of his dead father to offer some kind words to his son in a public forum. Other regional and opposition political parties, like the historic (and historically weak) Congress Party, shared deepfakes of celebrities and politicians endorsing preferred candidates or denouncing opponents (without those original speakers’ permission).
Modi himself spoke out against the trend late last year, but it only continued to intensify, to the point that various tech companies earned more than $50 million from political parties to produce A.I. images and videos, per Rest of World—which paid off quite well, considering how cheap it became to put this stuff out (“costs less than a cup of tea”). There was such a glut, in fact, that Modi and the BJP started labeling unfavorable videos as deepfakes even when they were very real. It was enough of a threat that they sent cops to arrest Congress Party workers who crafted deepfakes of Modi aides.
It’s obvious why the A.I. sludge was a core storyline in the year’s elections. Still, there’s reason to believe overall that the A.I. saturation was less effective, and more spammy and confusing—and that the key story as to why the BJP suffered a setback this round belongs to media dissenters who expertly weaponized their platforms in the face of oppression.
Independent news websites that took to digital channels to reach readers beyond newspaper circulation and TV broadcasts were hounded by the government to a frightening extent: Fahad Shah, editor in chief of the Kashmir Walla, was baselessly tossed in prison for “glorifying terrorism,” and the government shut down his publication shortly after. The founder of NewsClick was arrested in October under “terror” charges and freed only last month, after the country’s Supreme Court intervened. A similar incident occurred in 2022 with Mohammed Zubair, the co-founder of the fact-checking resource and frequent BJP poker Alt News.
Yet if the desired effect was to chill such alternate voices, the government’s actions had an opposite effect—boosting these journalists’ profiles and making them international free-speech heroes. That effect was magnified for India’s YouTubers, who’ve found ample success in reaching Indians across language and geographic divides as traditional media (often referred to as “Godi media”) loses credibility and audience share. Some famous YouTube pundits are even former TV newscasters.
None were more influential, perhaps, than Dhruv Rathee, an influencer-historian with 21 million subscribers and a relentless tendency to fact-check all the lies from and false media coverage of Modi and the BJP. Branding himself as an “independent” voice with digital savvy, whose videos gain traction across Instagram and X and WhatsApp, Rathee never softened his rhetoric against Modi and Co. throughout the election. While breaking down various scandals—party members paying off voters, Modi’s autobiographical embellishments, the national persecution of Delhi Chief Minister Arvind Kejriwal—Rathee used strong words like dictatorship and brainwashing, in a refreshing break from the pro-Modi mainstream media lines. When the election results came in, Indians took to X to point out how Rathee had been correct in his critiques of Modi, even as TV outlets kept insisting that the BJP would see an out-and-out victory this round.
Ultimately, despite the cheap deepfakes, Modi’s popularity, the violent nature of modern Indian politics, and a media apparatus in thrall to the BJP, digital journalists and influencers managed to counter the ruling party’s narrative in the places where it counted. As I noted back in 2021, a combination of post-COVID economic devastation and popular uprisings hinted at coming weaknesses for the Modi coalition, one that couldn’t just be papered away by Islamophobia and paeans to India’s greatness.
A lot of women, rural, and lower-caste voters felt that pinch, and they were receptive to the message from Rathee and his fellow YouTubers—that Modi, in fact, was not bringing them the prosperity he promised and was interested mainly in covering up for corrupt friends or stoking religious violence. There may be a lesson here for those worried about how deepfakes will influence another significant election this year: the United States’.