Technological advancement in Bangladesh has transformed society, and it is now opening a new chapter in electoral politics as well. With the proliferation of digital platforms and the easy availability of AI-powered tools, political campaigns are no longer confined to stages or leaflets. Social media is increasingly used to influence voter psychology through AI-generated fake videos, fabricated posts, and manufactured campaigns.
In the context of Bangladesh, especially with the 13th national parliamentary election approaching, this issue has become central to contemporary political discourse. AI-generated content is seen as a major threat, capable of misleading voters with false information and undermining the very foundations of democracy.
As the media gradually drifts away from the public and aligns with power, a new invisible and formidable player has entered the field – artificial intelligence. In journalism, politics, and electoral democracy alike, AI is no longer a prospect of the future; it is a stark reality of the present and the harbinger of potential dangers. The arrival of generative AI (e.g. ChatGPT) on the market has created as much excitement in the world of technology as it has profound implications for democratic politics. Writing news, editing, creating headlines, generating images and videos – all done with a single click using AI.
The question is: What is the cost of this for democracy? Jokers might say that news anchors were robotic anyway, with no room to go off-script. Yet, hidden within this jest is a frightening truth: when humans function like robots, the next step is to make robots human-like. AI has made this step even easier for those in state power.
Nicholas Diakopoulos, Professor of Communication at Northwestern University, rightly observes that AI is entering newsrooms, changing the nature of journalism. What he does not emphasise is that this transformation also shifts the balance of power. A 2019 global survey by LSE had predicted that AI’s impact would be structural and long-term. Now, standing at the threshold of 2026, we are seeing that structure rapidly turning into a political weapon.
The 13th national parliamentary election, scheduled for February 2026, could serve as a testing ground for this weapon. In Slovakia’s 2023 parliamentary elections, fake audio clips replicating the voice of an opposition leader were circulated on social media to mislead voters. In the 2024 New Hampshire primary, AI-generated robocalls misled voters with false information about election dates and procedures. According to research by Germany’s Konrad Adenauer Foundation (KAS), deepfake incidents have been reported in elections worldwide including the United States, Turkey, Slovakia, Argentina, Indonesia, India, Poland, Bulgaria, Taiwan, Zambia, and France.
It is now imperative to take coordinated and multidimensional steps to prevent AI misuse. Governments and regulatory authorities must formulate clear laws and policies that explicitly define AI-based fake information, deepfakes, and digital deception as crimes, with provisions for swift trials and exemplary punishments.
AI no longer merely writes news; it generates emotions, spreads hatred, and makes lies appear credible. Deepfake videos can make a candidate ‘say’ things they never did. AI-generated Facebook posts, memes, reels, and bot accounts can instantly poison public opinion. During elections, fake surveys, fabricated statements, altered images, and artificial trends misleading voters are no longer fiction.
When the media ceases to guard democracy and becomes a content factory for power, AI becomes its perfect machine. If journalism succumbs to cheap labour and politics becomes a game of data and algorithms, elections will remain a formality – votes will be cast, but opinion will vanish.
The real question is: will we control it, or will it control our politics, media, and democracy?
February’s election is not just a test of political power; it is the greatest moral test for Bangladesh’s digital democracy. If social and institutional resistance against AI, deepfakes, and algorithmic propaganda is not built now, in the future there may be humans at the polling booths but machines will make the decisions.
Shifting focus from street-level politics to theoretical discussion, we see that information flows in the rapidly evolving digital environment faster than ever in history. Through smartphones, social media, and AI-driven platforms, information is now created, disseminated, and transformed in real-time.
Opportunities are immense, but challenges are equally severe. Fake news, misinformation, disinformation, and algorithm-driven echo chambers severely compromise democratic decision-making. Christian Fuchs’ critical political economy perspective is particularly relevant here.
New media theorist and researcher Christian Fuchs, in his book Social Media: A Critical Introduction (2014), argues that seeing social media merely as participatory or democratic platforms is a false romanticism. Platforms like Facebook, YouTube, and X are essentially ‘corporate surveillance machines,’ where users simultaneously act as content producers and data laborers.
In Bangladesh, this is even clearer. A significant portion of young people create political opinions, memes, videos, and emotional content. But these are filtered by algorithms. Which post will go viral, which will be suppressed is determined by opaque platform algorithms, not citizen reasoning. Fuchs calls this ‘algorithmic power over public discourse.’
In Bangladesh, sensitive issues such as religion, the Liberation War, national security, or sovereignty often create echo chambers. Users are not accustomed to hearing dissenting views. Fuchs calls this ‘ideological enclosure of communication.’ The result is a shrinking of debate and diversity of opinion essential for democracy.
During elections, misleading surveys, fabricated statements, altered images, and artificial trends misleading voters are no longer fiction. The most frightening aspect remains – this propaganda has no face, no accountability. Questions cannot be asked, nor answers demanded.
In an emerging digital society like Bangladesh, if media literacy, ethical journalism, and algorithmic transparency are not ensured together, AI-driven new media will not strengthen democracy – it will silently erode it.
One of the most dangerous political consequences of AI-driven communication is the normalisation and accelerated spread of hate speech. In Bangladesh, hate speech on social media targeting religion, ethnicity, gender, political identity, or interpretations of the Liberation War is no longer merely personal animosity. It is the result of algorithm-driven visibility.
Even more concerning, in electoral politics and power struggles, hate speech has become a strategic weapon. In Bangladesh, online troll networks and bot accounts label specific groups or ideologies as ‘anti-state,’ ‘anti-religion,’ or ‘enemies.’ Fuchs terms this ‘digital nationalism and exclusionary communication,’ where hatred gains legitimacy in the name of national interest. The result is curtailed democratic debate, a culture of fear, and citizens conditioned into self-censoring silence.
In this context, combating hate speech is not just a matter of law enforcement; it is a combined struggle of media literacy, platform accountability, and political ethics. Without these, in digital Bangladesh, hate will become the most visible ‘public opinion.’
The writer is an Associate Professor, Communication and Journalism, University of Chittagong
