Two young Bangladeshi developers have unveiled an anonymous civic-tech initiative, known as ‘GhushSite’, engineered to record informal financial demands and illicit facilitation fees across government departments. Anchored in rigorous data privacy, the open portal enables members of the public to log instances of extortion without completing user registrations, providing email addresses, or submitting telephone numbers.
The civic ledger is the brainchild of Saif Kabir and Shahed Sharif, who conceived and deployed the platform in twenty-four hours after drawing inspiration from similar digital transparency projects in neighbouring India. The interface allows users to identify the government agency involved, the specific civic service requested, the general locality, the financial demand made, and the eventual outcome of the transaction. Crucially, contributors are barred from naming individual civil servants, submitting administrative serials, or uploading personal identifiers.
Kabir observed that whilst everyday citizens remain acutely aware of pervasive corruption, actionable data regarding standard extortion rates across various offices has stayed largely obscured. Individuals navigating routine public duties—ranging from passport renewals and driving licence issuances to land deed mutations—frequently encounter demands for under-the-table “speed money” without any baseline to determine whether such sums represent entrenched local tariffs or arbitrary inflation. By establishing an open database, the creators intend to map aggregate trends and highlight systemic vulnerabilities in public administration.
Within forty-eight hours of going live, the platform accumulated 458 unverified entries spanning all eight administrative divisions, representing roughly 24.5 million Bangladeshi Taka in reported demands. The regional breakdown indicates that the capital division of Dhaka generated the largest volume with 254 reports, followed by Chittagong with 98, Sylhet with 24, Rajshahi and Khulna with 22 apiece, Mymensingh with 16, Rangpur with 15, and Barishal recording seven entries.
Whilst the calculated mean across submissions stood at approximately 33,747 Taka, the creators noted that this figure was inflated by several exceptional claims in customs clearances, corporate taxation, and large-scale land administration, where alleged demands reached millions of Taka.
The submission architecture functions through a automated three-tier process. After an entry is drafted, it is processed by an automated language model acting as a content moderation filter. This algorithm strips out individual surnames, telephone digits, National Identity (NID) details, file reference numbers, and offensive prose before entries appear on the public feed. Kabir clarified that this filtering mechanism operates strictly as a data privacy safeguard rather than an investigative verification mechanism, ensuring that all published entries remain unverified citizen reports.
The founders maintain that the platform is not intended to substitute formal statutory anti-corruption bodies or act as a judicial tribunal. Instead, they position the initiative as a diagnostic tool. A single anecdotal claim of an extra-legal payment can be easily dismissed, but dozens of congruent testimonies concerning the same municipal office reveal persistent administrative anomalies. Self-funded by its creators and sustained by voluntary micro-donations, the ledger aims to provide researchers, investigative journalists, and policymakers with preliminary empirical insights whilst puncturing the longstanding culture of silence around low-level extortion.
