Independent Fact-Checking
Evidence-based verification of public claims, viral content, and misleading narratives.
Public Interest Information Infrastructure
Building a Trusted Information Ecosystem
AI-assisted verification with human editorial review, built for a more resilient information environment.
An initiative of Atlas AI Institute · Established 2025
Independent
Evidence-Based
Human-Centred
Transparent
Technology-Enabled
AI assists the process. Humans make the editorial decision.
What We Do
Strengthening the Information Ecosystem
AI-Assisted Verification
AI supports claim detection, evidence discovery, source aggregation, and analysis while humans retain editorial responsibility.
Information Integrity Research
Research into misinformation, disinformation, synthetic media, deepfakes, and emerging information threats.
Open Data & Knowledge
Structured claim records, datasets, research resources, and public-interest knowledge.
Media & Digital Literacy
Practical verification education for citizens, journalists, educators, and young people.
Public Verification Tools
Tools for citizens and researchers to investigate claims, sources, images, and other information.
The Verdict System
Five verdicts. Each one distinct — never interchangeable.
Every fact-check ends in one of five clearly defined, visually distinct verdicts. "Unverified" is never styled to look like "False."
True
Reliable evidence supports the claim.
False
Reliable evidence contradicts the claim.
Misleading
Contains elements of truth presented in a misleading way.
Partially True
Some elements accurate; others inaccurate or unsupported.
Unverified
Available evidence is insufficient for a reliable conclusion.
The Archive
A growing, evidence-based public record.
Every figure below reads from the live database. Nothing here is a fixed number in the page template.
Data currently being updated
Published Fact-Checks
Data currently being updated
Categories Covered
Data currently being updated
Public Corrections Logged
Bangla + English
Languages
Beyond Individual Fact-Checks
An information-integrity infrastructure, not a single feed.
Observatory
Monitoring misinformation, disinformation, and emerging narratives affecting Bangladesh, with verdict and topic trends drawn from the live dataset.
Research
Institutional research into AI-generated misinformation, synthetic media, platform governance, and Global South information integrity, led by Atlas AI Institute.
Dataset
A structured, multilingual dataset of public claims and editor-verified assessments. Candidate and verified records are always clearly distinguished.
Help Us Verify
Submit a claim, or fact-check where you already chat.
Submission does not guarantee publication or verification — selection is based on public-interest criteria.
Submit a Claim
Share a claim, source URL, screenshot, or context for our editorial team to review.
Telegram
@bdaifactcheck_bot — submit claims directly. AI gathers publicly available evidence; human editors remain responsible for published verdicts.
Facebook Messenger
FactCheckBD Messenger integration for claim submission. Image and video forensics may require the website's forensic tools.
An Initiative Of Atlas AI Institute
Institutional leadership. Independent editorial decisions.
Atlas AI Institute provides institutional, research, and methodological leadership. FactCheckBD is the applied, public-facing verification environment. Technology infrastructure does not control editorial verdicts.
Atlas AI Institute→
FactCheckBD→
Nagorik AI / SmartGen AI
Institutional & research leadership → Public-facing fact-checking → Technology & AI tooling infrastructure.
Editorial independence. No political party,
government agency, religious institution, or commercial sponsor
controls claim selection, investigation, verdicts, or publication.
There is no pre-publication political or government approval.
Evidence first. AI assists. Humans decide.
Help build a more resilient information environment for Bangladesh.