Misinformation and deepfakes in election season

Election season raises the stakes for false information. A misleading claim about a candidate, a voting date, a polling station, or an exit poll can travel faster than a correction, especially when voters are making up their minds. Misinformation is not one thing. It includes old videos shared with a new caption, edited clips, satire mistaken for news, and wholly invented claims. Disinformation is a narrower term, used for false material shared deliberately to mislead. Deepfakes add a further problem: audio or video that uses artificial intelligence to make a person appear to say or do something they did not.

The 2024 Lok Sabha election is a useful case study because the picture was more mixed than many headlines suggested. Fact-checkers who spoke after the election said older tactics still dominated: edited and manipulated videos, and old unrelated content reused with a new claim. For example, BOOM Fact Check reported that between 1 March and 31 May 2024 it published 258 election-related fact-checks in English, Bangla, and Hindi, of which 12 were AI-generated claims, including 3 deepfakes and 9 voice clones. That does not mean deepfakes were unimportant, but it warns against assuming every viral clip is AI-made — or that AI is the only threat. Some real videos were also wrongly called deepfakes, which can itself become a way to deny an inconvenient fact.

The Election Commission of India issued a communication on 6 May 2024 on the responsible and ethical use of social media, directing parties to avoid manipulated and deepfake content and to remove such content within three hours of coming to their notice. Enforcement, however, is difficult. A clip can be copied, re-uploaded, translated, and cropped before a platform or fact-checker acts. Supporters of AI in campaigns argue that the same tools can also help with translation, accessibility, and creative outreach, and that a clear label can separate creative use from deception. Critics argue that labels are easy to remove, that many voters will not notice them, and that even a debunked clip can leave a false impression.

For students, the lesson is to slow down at the moment of highest emotion. A claim that confirms a strong political feeling, or that outrages, deserves the most checking, not the fastest share.

Suggested picture (licence to confirm): The roots of 'fake news'-es.png — licence not confirmed, so the image is not embedded here.