Our partnership with Defence Holdings PLC continues to evolve, expanding beyond Defence and into UK national security and public safety.
As part of this next phase, Whitespace is working alongside Microsoft and Gloucestershire Constabulary to co-design and deliver an AI capability supporting the secure processing of taped and video interviews. The solution focuses on translation, transcription, summarisation and structured outputs, all delivered to a high standard of accuracy, assurance and operational trust.
Built for evidential integrity, not experimentation
In policing and national security environments, interview material is often central to investigations and intelligence workflows. Speed matters, but so does precision. Outputs must be reliable, repeatable and defensible, with confidence in how they were generated and how they can be used.
This work applies governed GenAI and agentic AI in a way that prioritises evidential integrity. Rather than treating AI as a generic productivity tool, the capability is being designed around real operational requirements, predefined formats and the assurance standards expected in sensitive law enforcement contexts.
Designed to scale across forces and use cases
While initially developed in collaboration with Gloucestershire Constabulary, the capability has clear potential to scale beyond a single force. The underlying approach is relevant across UK policing, international law enforcement, and wider intelligence and enforcement use cases where recorded interviews must be processed quickly without compromising trust or control.
This includes environments where language barriers, data volume and time pressure make manual processing increasingly difficult, but where uncontrolled automation is not an option.
Delivered on Collective
The solution is being built and deployed using Whitespace’s proprietary Collective Operating System. Collective enables the rapid development, adaptation and deployment of sovereign GenAI and agentic AI across different operational contexts, from cloud-based environments to more constrained or sensitive deployments.
By using Collective as the foundation, the team can iterate quickly while maintaining governance, auditability and control. This allows the capability to evolve in line with operational needs, rather than being locked into a single use case or deployment model.
A clear direction for 2026
This collaboration reflects Whitespace’s direction of travel for 2026: extending mission-ready AI beyond Defence and into national security and public safety, where real-world delivery, governance and assurance are non-negotiable.
It also demonstrates how partnerships across defence primes, technology providers and operational organisations can move AI from proof points into practical, deployable capability.
We’d like to thank the Microsoft and Gloucestershire Constabulary teams for their collaboration, and Defence Holdings for the continued partnership as we expand our joint work across Defence, national security and public safety.
