A police officer watching a crowded street sees only what is in front of them. A cyber police officer can potentially see patterns across thousands of phones, accounts, images, and videos.
Bangladesh is now developing the capacity to do much more of that.
On October 1, the country launched its new Cyber Police Unit with an approved structure of 4,592 posts. The unit is tasked with investigating cyber and financial crimes, analyzing digital evidence, monitoring social media and the dark web, and using AI-assisted video analytics and facial identification. Cyber desks are planned for police stations, alongside a 24/7 cyber support centre.
The case for expanding cyber policing is straightforward. Criminal activity increasingly leaves a digital trail, and police need the expertise and technology to follow it.
But a more capable cyber police force also raises a less comfortable question: Can the institutions governing these new powers develop as quickly as the powers themselves?
Digital policing is not merely about catching criminals more efficiently. It also concerns who can access information, how it is interpreted, and what happens when the state gets it wrong.
Bangladesh is not starting from scratch. The Criminal Investigation Department (CID) has operated its Cyber Police Centre since 2017 to deal with cybercrime and related digital investigations.
Since 2020, the Bangladesh Police’s Police Cyber Support for Women (PCSW) has provided legal and technological assistance to women facing cybercrime, including counselling and referrals to police stations, specialised units and victim-support centres. Broader reporting channels include police stations and the national emergency service, 999.
The challenge is not just creating more places for victims to report. It is making these channels work together.
A woman reporting image-based abuse, for example, should not have to know whether her case belongs with a local police station, a new cyber desk, the Cyber Police Centre, PCSW, or another specialized unit.
Her case should reach the right expertise without requiring her to navigate the bureaucracy herself.
That becomes more important as the new unit expands. The 4,592 figure represents approved posts, not a fully staffed force. Existing police personnel will initially fill positions before additional recruitment. The cyber desks and 24/7 support centre will also have to move from organisational plans to functioning services.
But the harder questions begin once the police have the data.
Who can access sensitive digital evidence? How long should it be retained? When can it be shared with another agency? How should investigators document the use of AI-assisted systems? And what happens when an automated identification or analysis is wrong?
A mistaken police conclusion about a digital identity, image, or online interaction can affect an investigation, a person’s reputation, or their freedom. As analytical tools become more powerful, scrutiny of their conclusions becomes more important.
Bangladesh’s Personal Data Protection Act 2026 establishes a broader framework for protecting personal data while allowing processing without consent where necessary for the prevention, detection, inquiry, investigation or prosecution of offences.
That is important for law enforcement. But a general data-protection framework cannot by itself determine how every police database, forensic system or AI-assisted tool should be operated.
Accountability matters especially because Bangladesh’s recent police oversight framework remains unsettled.
The 2025 Police Commission Ordinance included a mechanism for citizen complaints, but that framework is no longer in effect.
This creates a sequencing problem. It is relatively easy to acquire technology and create new police posts. It is harder to build durable institutions capable of scrutinizing how that technology is used.
Once databases, analytical systems, and specialized investigative practices become embedded, changing the rules around them becomes more difficult.
Bangladesh should build safeguards alongside capability. The new Cyber Police Unit should connect with existing victim-support and referral services rather than set up another isolated reporting channel.
Police should have clear operational rules governing access to, retention, and sharing of digital evidence, as well as procedures for documenting and challenging consequential errors in automated analysis. Citizens also need credible avenues to question misuse of digital investigative powers.
The future of cyber policing will not be determined only by what police can see. It will be determined by what they are allowed to do with what they see, and whether citizens have meaningful ways to challenge them when they get it wrong.
Ashfaq Ferdous is an academic at Southeast University, Bangladesh, and a former Assistant Superintendent of Police with Bangladesh Police.