As Bangladesh approaches its 2026 general election, the political battlefield has shifted dramatically. Campaigning is no longer confined to rallies, street posters, or televised debates. Increasingly, it unfolds in digital spaces-where algorithms dictate what citizens see, automated accounts amplify partisan narratives, and artificial intelligence fabricates entire realities. This transformation has profound implications for the democratic process. Understanding it requires the analytical lens of 'computational propaganda', a concept developed by Samuel C. Woolley and Philip N. Howard, in their seminal work, "Computational Propaganda" (2018). Their framework exposes how digital technologies can be systematically harnessed to manipulate public opinion, polarise debate, and shape perceptions of truth.
Computational propaganda is not merely misinformation or "fake news." It is the 'strategic use of algorithms, automation, and human coordination to disseminate misleading or manipulative content deliberately', unlike traditional propaganda, which relied on centralised messaging through print or broadcast media. Computational propaganda leverages scale, speed, and anonymity. Networks of bots, fake accounts, AI-generated content, and algorithmically optimised posts can create the illusion of consensus, amplify polarising narratives, and distort public understanding of reality. This is propaganda redesigned for the digital age, operating invisibly within the platforms that billions of users rely upon for information.
In Bangladesh, where more than 127 million voters are expected to participate in February 2026, the stakes could hardly be higher. Social media platforms-particularly Facebook, TikTok, Instagram, YouTube, and X (formerly Twitter)-have become central to political discourse. Yet alongside their democratic potential, these platforms have emerged as vectors for disinformation and targeted manipulation. Fact-checking organisations have documented hundreds of instances of political misinformation in the first weeks of 2026 alone, ranging from AI-generated videos to deepfakes that distort statements by party leaders or fabricate endorsements by ordinary citizens.
Woolley and Howard's framework emphasises that computational propaganda is 'systemic', operating not only through the content itself but also through the mechanisms of algorithmic amplification, networked coordination, and human-machine interaction.
“In Bangladesh, where more than 127 million voters are expected to
participate in February 2026, the stakes could hardly be higher. Social
media platforms-particularly Facebook, TikTok, Instagram, YouTube, and X
(formerly Twitter)-have become central to political discourse”
AI-generated content has proliferated across social media. Deepfakes, synthetic videos, and AI-created avatars simulate political endorsements, campaign messaging, or citizen support for particular candidates. These materials are often visually and aurally convincing, making it difficult for the average voter to distinguish real from fake. Such "synthetic authenticity" is a hallmark of computational propaganda: content that feels real but is entirely fabricated. In January 2026, the non-profit fact-checking organization Rumor Scanner identified more than 500 instances of AI-driven political misinformation circulating online, illustrating both the scale and sophistication of these efforts.
Algorithmic amplification further magnifies these effects. Social media platforms are designed to promote content that generates engagement, regardless of its truthfulness. Posts that are sensational, emotionally charged, or polarising are more likely to trend, be recommended, and appear in users' feeds. In Bangladesh, Facebook and X have been the primary arenas for such amplification, although Instagram, TikTok, and YouTube are increasingly leveraged to target younger, urban audiences. The interplay between algorithmic curation and human behavioural biases-such as confirmation bias-creates digital echo chambers in which misinformation spreads rapidly and entrenches itself.
Coordinated networks of humans and bots intensify the impact. Political operatives, influencers, and paid promoters seed, share, and curate content, while bot networks systematically like, retweet, and comment to simulate widespread consensus or controversy. In some cases, bot amplification is combined with AI-generated content to produce entire campaigns that appear to be organic citizen movements. Woolley and Howard describe this human-machine synergy as central to modern propaganda: machines optimise for reach and virality, while humans supply creativity, strategy, and context-sensitive targeting.
The consequences for democratic integrity are profound. Computational propaganda does not merely distort information; it 'reshapes the informational environment itself', making it difficult for voters to discern authentic political discourse. Exposure to manipulated content erodes trust not only in individual leaders or parties but also in media institutions and the electoral process itself. When social consensus can be engineered digitally, elections risk being decided not solely by citizen choice but by the invisible mechanics of algorithmic persuasion.
Bangladesh's 2026 election has already witnessed these dynamics. Beyond X account hacking, there have been reports of coordinated disinformation campaigns on WhatsApp groups, TikTok trends promoting partisan content, and Facebook pages propagating doctored images and videos of political leaders. This multi-platform strategy illustrates the adaptability of computational propaganda: propagandists do not rely on a single medium but exploit multiple channels to reach diverse demographics, reinforce narratives, and sow confusion.
The speed and scale of manipulation compound the challenge. Corrective measures-fact-checking, content moderation, and public advisories-often lag behind the initial dissemination of false content. By the time a post is flagged or removed, millions of users may have already seen, shared, and internalised the message. This temporal advantage amplifies the impact of propaganda, creating conditions in which perception often outweighs reality.
Addressing these challenges requires more than reactive fact-checking. Regulatory frameworks that monitor algorithmic manipulation, stronger safeguards against account hacking, and public digital literacy initiatives are essential. The Bangladesh Election Commission has taken preliminary steps, requiring candidates to declare social media accounts and collaborating with platforms to detect disinformation. However, effective enforcement will be critical, as propagandists continuously innovate and exploit technological loopholes.
Ultimately, computational propaganda in the Bangladesh election is a 'symptom of a broader digital transformation in political communication'. It demonstrates that while technology can democratize information and participation, it can equally be weaponised to manipulate opinion at scale. Woolley and Howard's framework reminds us that the danger is not merely the presence of falsehoods but the organised, systemic use of computation to shape political realities.
For Bangladesh, the 2026 election will test not only political leadership but also the resilience of democratic institutions, the integrity of information ecosystems, and the capacity of citizens to navigate a complex digital landscape. Safeguarding electoral integrity will require vigilance, innovation, and a commitment to ensuring that truth-and not just the digital signal-guides the nation's democratic choices.
The writer is a researcher and development professional
