AI fake videos: A new threat to India’s democracy and society

The Election Commission of India (ECI) recently issued a strong notice to political parties, warning them to take down fake and AI-generated videos within three hours of being reported. This directive came after a series of deepfake videos surfaced during election campaigns, some even showing well-known film stars apparently endorsing political parties. These incidents underline a disturbing truth: India is entering an age where technology can distort reality faster than the truth can correct it.

A democracy under digital siege

The heart of democracy lies in informed choice. But when voters see realistic fake videos of leaders saying or doing things they never did, that choice is corrupted. Deepfakes use AI tools to clone voices, replicate faces, and create hyper-realistic scenes. A single manipulated clip can mislead millions, especially in a country where digital literacy remains uneven.

During the 2024 Lok Sabha campaign, a fake video of actor Aamir Khan appeared online, showing him appealing to people to vote for a particular party. Soon after, another video of Ranveer Singh was circulated in a similar fashion. Both were confirmed to be AI-generated, but not before they had gone viral, shared widely on WhatsApp and social media. The damage was already done, millions had seen them, many believed them, and a shadow of doubt lingered even after fact-checking.

The ECI rightly warned that the use of “AI-based tools that distort information or spread misinformation lowers the standards of electioneering.” Democracy suffers when the battlefield of ideas is replaced by a theatre of deception.

Beyond politics: the social and economic ripple

Fake videos do not just distort politics they also poison society. Deepfakes can impersonate private citizens, defame women, or inflame communal hatred. A manipulated video shared in the wrong group can spark violence, as seen earlier with WhatsApp rumors leading to mob lynchings across rural India. Visual evidence, even when false, often triggers emotional reactions before logic sets in.

Economically too, the risks are real. Imagine a fake video claiming that a major bank is collapsing or a corporate CEO is resigning. Stock markets can crash within hours, and investor confidence can evaporate before the truth emerges. In a world driven by perception and speed, misinformation becomes a hidden tax on trust and stability.

A crisis of trust

The rise of deepfakes creates something even more dangerous than misinformation disbelief. When people begin to suspect that every video might be fake, they stop trusting anything at all. This “liar’s dividend” lets wrongdoers escape accountability by dismissing real footage as fake. The result is a society where truth becomes negotiable.

In India’s vibrant democracy, where millions get their news through short clips and forwarded videos, this crisis of trust could be catastrophic. It could silence genuine voices, discourage whistle-blowers, and make citizens doubt even verified sources. For a society already divided by language, class, and politics, deepfakes add a new layer of confusion and cynicism.

Technology outpacing law

The Election Commission’s directive is a crucial step, but it’s only the beginning. India’s legal system is still catching up with AI-driven manipulation. While the Information Technology Rules 2021 and the Representation of the People Act 1951 provide a broad framework, they were not written for a world where anyone with a laptop can clone a face or voice in minutes.

Currently, it is difficult to trace the original creator of a fake video. Platforms can be slow to act, and political parties often use “proxy accounts” to circulate questionable content while maintaining plausible deniability. In many cases, by the time a video is verified as fake, it has already achieved its intended impact shaping opinions and sowing division.

Damage already visible

Several recent cases highlight the scale of the threat. Apart from the celebrity deepfakes, fake videos showing fabricated communal violence or false claims about government policies have circulated widely. In Uttar Pradesh’s Meerut region, a video falsely claiming drone attacks led to panic and multiple arrests. Even though it wasn’t AI-generated, it revealed how visual misinformation can disrupt public order and waste administrative resources.

In rural areas, research shows that forwarded videos, many misleading or out of context, dominate WhatsApp conversations. Fact-checking rarely reaches these audiences, especially in regional languages. The poor and less-educated citizens thus become the most vulnerable targets of digital deception.

What must be done

Combating AI-driven misinformation in India’s elections demands a layered approach: enforce mandatory disclaimers on all AI-generated campaign videos, require swift content removal as per ECI directives, and compel platforms to strengthen detection and watermarking of synthetic media. Public education in regional languages is critical so citizens can spot deepfake clues like mismatched lips or altered voices. Political parties should adopt an ethics charter against digital deception, while laws must sharpen penalties for election-related impersonation. Collaborative research by academia and civil society will further improve defences against deepfakes. However, technology isn’t enough citizens’ vigilance, journalistic verification, and political responsibility form the real barrier. With over 900 million internet users, India’s battle with deepfakes will influence the global south’s response. The true threat is not just the prevalence of fake content but the erosion of trust, which, if left unchecked, risks undermining democracy’s very foundation.