We are implementing our new website. Some features may not work properly.
DataResearchDataXplainer ↗About
← Research
Information integrity

The Disinformation Election

What the evidence shows about the forms election disinformation took in Bangladesh in 2026, how it spread through the observed Facebook ecosystem, and who it targeted.

Research
Dr. Saikat Ghosh
Adviser
Dr. Ahmed Abidur Razzaque Khan
Period analysed
December 2025 – February 2026
Evidence
2,685 Facebook posts
Research question

What did election disinformation look like, how did it spread, and who did it target?

77Accounts
2,685Posts analysed
22×Fake-news surge
955K interactionsMost viral post
What the evidence shows

Key signals from the research

01

Fake-news activity accelerated before election day

Posting by the observed fake-news pages rose from an average of 6 posts per day in mid-December to 134 posts on 11 February 2026, the day before the election.

02

Journalists were a major target

The study identified 566 posts attacking editors and journalists, framing them as agents, operatives or enemies.

03

Anti-India narratives drew unusually high engagement

Anti-India stories generated 6.8 times more reaction than election content in the observed dataset.

04

The ecosystem showed strong thematic overlap

Eight influencer pairs had identical topic coverage in the network analysis, alongside a wider network of highly similar thematic patterns.

Method

How the research was built

Sources37 political influencers and 40 pages posing as news organisations.
AnalysisContent analysis, engagement statistics and network analysis.
Collection windowInfluencer posts: 10 Dec 2025–23 Jan 2026; fake-news posts: 11 Dec 2025–12 Feb 2026.