In the frenzied digital arena ahead of Bangladesh’s national election, a new weapon has emerged, one that trades in mockery instead of missiles. It is the coordinated “haha” reaction—a simple Facebook emoji deployed by armies of fake accounts to ridicule, suppress, and manipulate public perception. An investigation has uncovered a vast, commercially available network of bot profiles and “click farms” that are being hired across the political spectrum to artificially amplify support for allies and orchestrate waves of derision against opponents. What appears to be spontaneous public scorn or organic popularity is often a carefully purchased illusion, cheaply bought and swiftly delivered, threatening to distort the democratic process by manufacturing a reality that does not exist. This digital subterfuge exploits the very algorithms that govern what we see online, pushing genuine concerns and political speech out of view under a avalanche of manufactured laughter.
How Does a Simple ‘Haha’ Become a Political Weapon?
The mechanism is deceptively simple yet profoundly effective. When a political actor, activist, or even an ordinary citizen posts something online, coordinated clusters of accounts swarm the post with “haha” reactions. This is not mere trolling for its own sake. Internal Meta documents previously leaked as the “Facebook Papers” revealed that platforms like Facebook can automatically reduce the visibility of posts that show signs of being trolled or engaged with negatively. Therefore, a flood of “hahas” does more than mock; it actively signals to the algorithm that the content is objectionable, potentially ensuring fewer people see it in their feeds. This turns a feature of social interaction into a tool for digital suppression. An investigation analyzing 263 political posts found clear evidence of these organized “digital mobs.” At least four distinct, politically-aligned clusters were identified—two supporting Jamaat-e-Islami, one for the Awami League, and one for the Bangladesh Nationalist Party (BNP)—moving in unison to attack opposing camps. Their primary target is often not just rival parties, but also independent media outlets and journalists, aiming to discredit factual reporting and amplify partisan narratives.
Who is Selling This Artificial Engagement and How Does It Work?
The infrastructure enabling this manipulation is a shadowy but straightforward marketplace. Numerous Facebook pages and online portals, known as “click farms,” openly advertise and sell social media engagements. Posing as a client, investigators easily purchased 30,000 reactions—both “love” and “haha”—for dummy posts over just two days. The process required no verification; only a post link and a mobile wallet payment. The sellers, with names like Sociafy and Tech Dream, operate by maintaining “servers” or collections of thousands of fake Facebook profiles. These accounts are often bought for a pittance, sometimes using AI-generated profile pictures and foreign-sounding names with no plausible link to Bangladesh. Reactions from these international bot accounts can be delivered astonishingly fast—thousands within minutes. “Domestic” bots, with Bangladeshi-sounding names and slightly more detailed profiles, are more expensive and slower, but are considered less likely to be detected and deleted by platform moderators. Crucially, the investigation found that the same pool of bot profiles purchased for testing was also actively engaged in real political attacks, proving a direct commercial pipeline from click farms to the political front lines.
Which Political Actors Are Using These Bots and To What End?
The use of these services is disturbingly widespread and bipartisan. The investigation found bot accounts involved in two key ways: for attack and for artificial inflation. On the attack side, coordinated “haha” clusters affiliated with different parties target opponents. For instance, pro-Jamaat clusters were found to be the most aggressive, frequently targeting BNP and left-leaning content, while an Awami League-aligned cluster focused its mockery on Jamaat and the National Citizen Party. Perhaps more alarmingly, these services are also used to fabricate a facade of support. The investigation traced at least 547 identified bot accounts to the official pages of six parliamentary candidates from parties including the BNP, Jamaat, and Islami Andolan Bangladesh. These candidates had thousands of bot followers, many sourced from the same click farms that sell reactions. In a telling revelation, click farm owners confirmed that politicians from rival parties often buy from the same suppliers, sometimes even targeting the same post with opposite reactions—one buying “hahas” to mock it, the other buying “loves” to boost it. This illustrates a cynical, transactional relationship where digital mercenaries serve all sides, undermining authentic political discourse.
What Does This Mean for Democracy and Can It Be Stopped?
This ecosystem of artificial engagement presents a clear danger to electoral integrity. It allows well-funded actors to create a false sense of consensus, marginalize opposing voices, and drown out reputable journalism with orchestrated ridicule. It violates the terms of service of social media platforms, which explicitly prohibit the artificial inflation of engagement. However, as the investigation shows, Meta’s detection systems appear to be failing to curb these operations in the Bangladeshi context. The solution is multi-layered and challenging. Platforms must significantly improve their ability to identify and dismantle coordinated inauthentic behavior, especially in non-Western language contexts. Electoral authorities and civil society need to build greater digital literacy, helping the public identify signs of manipulation. Ultimately, the market exists because there is demand. As long as political actors believe that manufactured mockery and fake followers are effective tools, they will continue to invest in them. The 2026 Bangladeshi election has become a case study in how cheap, accessible technology can be weaponized to skew the democratic battlefield, replacing debate with a purchased laugh and substituting genuine support with a battalion of digital ghosts. The lasting impact may be an eroded trust not just in political opponents, but in the very online spaces where modern democracy is increasingly debated and decided.
