Rising fake videos, disinformation threaten elections, EC’s capacity questioned

The misuse of artificial intelligence (AI) is escalating ahead of Bangladesh’s parliamentary elections, with deepfake videos and manipulated images spreading rapidly on social media to discredit political opponents and sway public opinion. 

The Election Commission (EC) acknowledges this as a threat to fair elections but has made little progress in addressing it, despite including AI misuse in the Representation of the People Order (RPO) amendment.

Experts highlight that approximately 80 million Bangladeshis use the internet, with 70 percent active on social media, many of whom do not verify information before sharing it. 

This vulnerability amplifies the impact of fake content, potentially affecting millions. 

Politicians and analysts warn that without effective measures, election transparency and credibility could be severely undermined.

Cyber security trainer Arif Moinuddin stressed the need for modern technology to detect deepfakes, alongside financial support for media and fact-checking platforms, and increased technological literacy among the public.

A DismissLab report noted a 17 percent surge in fake information during May and June, with 44 percent tied to politics. 

AI-generated content, including fake videos targeting female leaders, is prevalent on platforms like YouTube, TikTok, WhatsApp, X, and Facebook. 

Of 176 fake graphic cards using media logos, 85 percent were politically motivated. 

Miraj Ahmed Chowdhury of DismissLab suggested that regulated AI use in campaigns could benefit financially weaker parties.

Globally, similar issues have emerged, with deepfake campaigns targeting Joe Biden in the US, over 100 candidates in India’s Lok Sabha elections, and Romania’s cancelled presidential election due to AI interference. 

The EU, US, UAE, India, and China have implemented AI regulations, but Bangladesh lacks comprehensive laws, limiting the EC and law enforcement’s ability to act.

Political parties, including the Jatiya Nagorik Party, BNP, and Jamaat-e-Islami, have urged the EC to collaborate with the government to curb AI misuse, warning of its potential to disrupt elections and discourage women from politics.

The EC’s revised code of conduct bans harmful AI use, including hate speech and false information, with penalties raised from Tk50,000 to Tk150,000 and possible candidacy cancellation. 

However, EC official Abul Fazal Md Sanaullah admitted challenges in regulating social media AI use, while EC Secretary Akhtar Ahmed mentioned plans for a monitoring team. 

Experts like Sumon Ahmed Sabir and fact-checker Kadruddin Shishir question the EC’s capacity, emphasising the need for a dedicated cybersecurity unit and collaboration with fact-checking organisations to counter disinformation effectively.