Bangladesh govt warns officials not to feed confidential data into AI tools over privacy risks
In a Nutshell:
- The Cabinet Division issued instructions on AI use in government offices on 28 July.
- The directive was sent to all senior secretaries and secretaries through the Cabinet Division’s e-governance 2 branch.
- The government said careless AI use could expose confidential information to unauthorised parties.
Why it matters
Government offices hold some of the most sensitive information in the country, and a major leak could damage public trust and disrupt important state functions. A recent case in Mexico this February showed how AI tools can be misused in cyberattacks targeting government systems, with researchers saying hackers used AI to help steal tax, voter and employee data. As Bangladesh begins using AI more in official work, authorities need clear safeguards in place before these tools become a regular part of government operations.
What we think
Historically, within our government framework, strict directives of this nature are rarely purely proactive; they are typically reactive measures implemented only after a significant data leak or systemic failure has already occurred. While this new AI guideline could be viewed as a precautionary step, our country's ongoing struggles with digital security and frequent data breaches suggest otherwise. It is highly probable that an undisclosed incident has already triggered this response. Although issuing such a directive indicates the government is taking the threat seriously, true commitment remains to be seen. Until we witness the execution of concrete and effective security measures, this remains merely a warning on paper.
