Unlabelled AI Videos Raise Voter Confusion Ahead of Polls

Representational image

As the election approaches, social media platforms are flooded with campaign content promoting political parties, much of it generated using artificial intelligence (AI).

However, most of these AI-created videos are not labelled as such, raising concerns about voter confusion.

Between 1 and 15 January, fact-checking organisation Dismislab identified around 800 AI-generated political videos, publishing their analysis on Saturday. It found that 60% of the AI videos on Facebook carried no indication that they were AI-generated.

During this period, 576 AI videos were posted from 21 Facebook pages. Similar content appeared on seven YouTube channels and two TikTok accounts. Among 181 videos on YouTube, 94.48% had no AI label, while none of the 50 videos on TikTok were labelled.

Dismislab also highlighted inconsistencies on Facebook: the “AI info” label appears on the main app, but not on Facebook Lite or desktop platforms.

Risk of Influencing Voters

Global research and Dismislab’s findings suggest that AI videos can influence voter opinions. Many viewers believed the videos depicted real events.

In one AI video promoting an Islamist party, a woman describes the party’s electoral symbol as a “symbol of justice.” Commenters praised the message, with one writing, “Masha’Allah, little sister. May Allah grant you understanding of the deen.”

Another video supporting a different party prompted comments urging people to back a specific candidate as the next prime minister. Besides promotion, some AI videos also spread false information or attack opponents.

Suman Rahman, Dean of the School of Social Sciences at the University of Liberal Arts Bangladesh, said such videos inevitably shape voter perceptions. “People tend to believe content that aligns with their political views. Negative information about opponents is often accepted without verification,” he said.

Legal Implications

Under the Representation of the People Order, using AI to create or spread false, misleading, biased, or defamatory content intended to mislead voters is prohibited. The election code of conduct also bans AI-generated misinformation.

Quoting Election Commission Public Relations Officer Md Ruhul Amin Mallik, Dismislab noted that spreading false information would be considered a dishonest intent.

Lawyer Jyotirmoy Barua urged the Election Commission to issue clear directives requiring all AI-generated content to be labelled, to protect voters from confusion.