Sometime in February, a website appeared claiming to run Bangladesh’s new ‘Family Card’ welfare programme online. The URL read ‘prothom-alo.online.’ The design, the layout, the trust it borrowed – belonged to Prothom Alo. The newspaper had never built it, and it wasn’t the first time someone had done this to them. That theft set the pattern for what followed. That’s the pattern of misinformation in Bangladesh right now. It doesn’t invent a source; it steals one, and the theft works best on readers with the least means to check it.
The photocard has been a normal part of how Bangladeshi outlets distribute news on Facebook for years. Lately, however, it has become the favourite costume of people with no news to report. A written rumour still asks a reader to process and weigh a sentence; a photocard substitutes a logo and headline for that work. For a reader with little practice separating a real outlet from an imitation, that substitution is the whole trick – nothing is left to verify because nothing was ever really read.
In August 2025, a card put a fabricated restraint appeal about Dhanmondi 32 into BNP leader Nazrul Islam Khan’s mouth. During the Sylhet stone looting, two fake cards wearing newspaper logos claimed a chairman said the stones were only being washed, and that a Jamaat leader had built a palace from the proceeds.
Outlets whose identities got copied during the election period included Kaler Kantho, DBC News, Jamuna TV, Channel24, Jugantor, Amar Desh, BBC Bangla and Prothom Alo – a spread wide enough that no single newsroom could monitor its own name across all of them. Prothom Alo’s own corrections history shows what that looks like from the inside: a fabricated Peter Haas quote about US satisfaction with the election, a fake 5,000-taka Eid promise under a doctored logo.
If a trusted masthead can fool readers who follow the news closely enough to catch a correction later, the same trick against a reader with fewer resources to double-check leaves little room for doubt – the fake gets believed and passed along. That is the engagement engine underneath the whole trend: a claim wearing a trusted logo, touching politics or sexual violence, spreads because it is felt before it is questioned – and a reader with less practice questioning headlines has the least defence against that design.
A FactWatch review of the 2026 election period found 72 separate cases of AI-manipulated content, 28 percent of it fake quotes stitched into photocards carrying real media logos – concluding that the volume alone overwhelmed fact-checking capacity. A correction or fact-check reaches only the audience already following the fact-checker’s page – not the reader who already shared, believed and moved on, precisely the reader this format is built to catch.
A masthead-mimicking card – built to be mistaken for Prothom Alo or Amar Desh – is easy to name as fake once someone checks the real outlet’s feed. A new variant skips that step: a generic ‘breaking news’ layout without any outlet’s name, headline typography, or a stock photo – just enough visual grammar of journalism to read as credible, with nothing left to disprove because nothing specific was impersonated.
A related trick surfaced repeatedly around the election: a parody account’s joke, stripped of its satire branding and recirculated as a real statement – a fabricated declaration, never posted on any verified account, moving through timelines with the label quietly removed. Nothing here impersonates a named newsroom, so the fastest debunking route – ‘the outlet says it never published this’ – does not apply.
What is left is slower, source-by-source verification, against a volume fact-checkers already struggled to match. For a reader with fewer tools to verify in the first place, this unbranded form removes the last visible clue that something might be wrong.
The law offers little. No provision found targets counterfeit news branding. The 2025 ordinance dropped provisions on impersonation and false publication, and nine sections in all. The proposed Section 26A would punish rumour and disinformation with up to ten years. However, critics say ‘rumour’ means unverified information, which describes much of breaking journalism. Therefore, the risk falls on reporters and ordinary sharers, while anonymous page operators stay hard to reach.
Fact-checking and rating, however diligent, treat each card as a single claim to disprove after the fact; they cannot, alone, outrun a format built to be believed before anyone has time to check. Media literacy is the answer – building the habit of doubt takes sustained effort. Until that habit outpaces the format exploiting its absence, the newsroom whose name is on the card will keep being the last to know it was used.
The views expressed in this article are solely those of the author
The writer is a Research Coordinator, FactWatch
