Helping police redact sensitive information in media files
Building evidence for the creation of an efficient and future-facing system that can be rolled out nationally.
Police and law enforcement often need to share material with courts or other parties whilst also protecting sensitive details such as identities and licence plates for data security or operational security reasons.
However, there is currently very limited use of auto-redaction technology across policing and the wider criminal justice system for digital media, audio, and video files - including body-worn video footage and digital forensic evidence - and the policing minister has made development of solutions which could be rolled out nationally a key priority.
The Home Office funded the Accelerated Capability Environment (ACE) to carry out a market review of existing multimedia redaction tools and build an evidence case into how state-of-the-art technology, including the use of artificial intelligence (AI) and machine learning, could make significant efficiency savings.
This would then be used to accelerate the development and widespread adoption of this technology nationally, creating user efficiencies as well as ensuring that police can confidently share information with other organisations as needed.
Understanding existing capability
To build this evidence foundation, ACE first engaged a business analyst to understand more about current redaction tools, software and techniques being used or developed locally by individual forces and shortlist the best performers.
Suppliers from our Vivace community then designed a suite of tests to assess the suitability of the shortlisted systems and examine their capabilities using synthetic data provided by Surrey Police. Results highlighted both uses and limits, with a report compiled on the systems recommended for further exploration.
In the next phase of this fast-paced commission, six candidates were invited to demonstrate their tools to key senior stakeholders within the Office of the Policing Chief Scientific Adviser, the Home Office, and policing, running live redaction on the synthetic data to showcase the extent of the automated redaction potential as well as the art of the possible.
The aim of this phase was to understand and document the capabilities of available tooling, cross reference them against user requirements, and conduct a gap analysis to identify where there are no or limited capabilities.
The Home Office is using the reports created by ACE to decide the next steps that will deliver the best AI-powered automated redaction tools into policing at the earliest opportunity.