Unlabelled AI-generated videos spread in election campaigns, reports Dismislab
Between 1 and 15 January, Dismislab documented more than 800 AI-generated videos related to election campaigns across Facebook, YouTube and TikTok, including duplicate counts.
Unlabelled AI-generated videos are circulating widely on social media as part of election campaigning in Bangladesh, raising concerns over platform enforcement and voter awareness, according to independent fact-checking platform Dismislab.
In one Facebook video posted on 4 January, an elderly man is shown in a busy marketplace saying he will vote for the Sheaf of Paddy, the electoral symbol of the Bangladesh Nationalist Party (BNP), citing promises of family, agriculture and health cards.
In another video posted a day later, a woman standing in a rural setting praises the Scales, the electoral symbol of Jamaat-e-Islami, describing it as a symbol of justice and fairness.
According to Dismislab, closer inspection reveals inconsistencies in both videos. Each clip contains frequent scene cuts, with a new scene appearing roughly every eight seconds.
Dismislab reports that both videos are AI-generated and are being circulated on social media as part of political campaigning ahead of the 13th National Parliamentary Election scheduled for 12 February.
While the use of AI in political campaigns is not prohibited, none of the videos contain an AI label, which violates Meta's transparency policy.
Between 1 and 15 January, Dismislab documented more than 800 AI-generated videos related to election campaigns across Facebook, YouTube and TikTok, including duplicate counts. It found that 60% of AI-related content on Facebook carried no AI disclaimer.
The situation was similar on other platforms. Of the 21 Facebook pages identified, seven posted the same videos on YouTube and two on TikTok.
A review of 181 videos from the seven YouTube channels showed that 94.48% had no AI label. None of the 50 videos posted on TikTok contained an AI disclaimer, according to Dismislab.
On Facebook, where 576 AI-generated videos were posted from 21 pages, Dismislab found that the "AI info" disclaimer is visible only on the primary Facebook app. When the same videos are viewed on Facebook Lite or on a computer, the label does not appear.
Dismislab noted that in Bangladesh, where digital literacy remains low, many users mistake AI-generated content for authentic information.
It also reported that the Bangladesh Election Commission has expressed concerns over the spread of AI content on at least two occasions, first on 12 December and again on 20 January, two days before official electioneering began.
Dismislab said its findings, alongside global studies, show that AI-generated content can influence voter opinions and create an uneven electoral environment, particularly when such material is circulated without disclosure.
Experts cited by Dismislab warned that voters are particularly vulnerable to AI-generated content because many tend to trust information that aligns with their existing political beliefs.
During elections, this tendency intensifies, increasing the likelihood that such content will be accepted and shared without verification, especially when it involves negative claims about political opponents.
According to Dismislab, the AI-generated videos feature a wide range of fabricated characters, including supposed police officers, political figures and ordinary citizens, who appear to endorse parties and promote favourable narratives.
The analysis found that content supporting Jamaat-e-Islami often used false endorsements by "official figures" and common people, while several BNP-linked videos used AI-generated children to seek votes.
Both sides were also found to be using AI-generated videos to discredit opponents.
Content favouring Jamaat frequently portrayed the BNP as extortionists or deceptive, while counter-campaigns targeting Jamaat used similar tactics to depict the party as "anti-state", Dismislab reported.
