The country has never had such copious information on women's safety as it has now. Police log the complaints, initiates the investigation process, and companies report redressal numbers. However, when it comes to responding a woman facing a threat, the reaction time is the same as that a decade ago. This is the impediment institutions need to address.

Based on the concept of intelligence pyramid that emphasises on the ‘intelligence-to-action gap’, challenge here is not collection of data, but conversion of the available data from mere numbers to actionable points before harm is multiplicative. And the gap of the unestablished connection between the information at frontline and that with the decision-maker acts that actually costs organisations/systems. While data is a signal, information is a signal organised and intelligence is the information that goes to a decision and is given to somebody who can act upon it. Collection is not intelligence. Every institutional failure to protect a woman is, underneath, an intelligence failure. Beneath each failure of protecting a woman is an intelligence failure. The right information existed somewhere, it just needed to be found. However, it reached the wrong person, in the wrong form, at the wrong time or it didn't reach anyone who had the authority to act.
A first information report is stored on police database, a harassment report is in an HR database, a distress call in a helpline log. All of these systems operate in silos without communicating with one another. A woman's safety is determined by whether these systems intersect, not on whether any institution designed them to do so.
{{/usCountry}}A first information report is stored on police database, a harassment report is in an HR database, a distress call in a helpline log. All of these systems operate in silos without communicating with one another. A woman's safety is determined by whether these systems intersect, not on whether any institution designed them to do so.
{{/usCountry}}Almost always, when a decision must be made, national policy calls for reporting upward to a regulator or the government, but never sideways amongst the institutions. The numbers of harassment have to be disclosed under the Companies (Accounts) Second Amendment Rules, 2025 by the ministry of corporate affairs. Platform transparency reports are needed to regulate digital. Unfortunately, nothing deters the predator from changing his job six months after his last complaint.
The picture is no different within a single organisation. Case file is in the custody of HR. Internal report of an abuse message rest with the IT department. The settlement of a prior complaint has been kept quiet by Legal. Three departments, three pieces of information around the same pattern, but until a formal complaint is made, these pieces don't come together. This lacuna makes it easy for a serial harasser to move between organisations. The HR proceeded with his quiet exit and complaint didn’t get the desired attention.
The most discernible dividing line is between the raw signal as it enters the front line and the rest of the signal. A police station has jurisdiction over the address of the complainant, but the server address from which images were post, the platform's compliance team or the address of the state where the accused is living at present are out of police’s purview. A content moderator could delete one image, not the network that distributes it. A deepfake could be circulating for weeks, for lack of a nodal agency that has both power and responsibility to act immediately.
These silos are no coincidence. The reasons for the existence of jurisdictional boundaries between state and central agencies are because policing is state-specific. Most of the laws are primitive, drafted before digital harm was a reality. These reasons collectively ensure a gap what a frontline officer knows, what an institution decides and what policy permits.
Better intelligence does not mean collecting large amount of data but creating systems that predict a pattern rather than waiting for complaints resulting in early signs of a problem. It means creating legally sound channels for verified conduct for transfer of information (not raw data), between the departments and the agencies once the evidence reaches a threshold.
It means imposing a response timeframe on each escalation because complaint without set response time is anything but. It means architects and designers considering privacy at the initial stages. There is no conflict between privacy and sharing of information for the purpose of intelligence. Good design isn't an overt disclosure of the individual and it also conforms to a pattern.
There is more data on women's safety than ever in India. Unfortunately, the response is not any quicker. This doesn't mean they're at fault in the collection. It's a decision they make, purposely, not to link what already has been. The fix is not a new dashboard or annual disclosure but agreements amongst institutions to interpret a signal, and mandate timelines that turn a flag into a deadline. Until institutions work on optimal utilisation of the available information, every failure will keep looking like a mystery, when the information existed right by the side, guarded by someone with no commitment to share it.
(The views expressed are personal)
This article is authored by Sagarika Chakraborty, CEO, India & Gulf and Garry Singh, president, IIRIS Consulting.