Skip to content

Interoperability in Justice and Public Safety Forum – Fifth session write-up

8 October 20254 min read
News
Interoperability in Justice and Public Safety Forum  – Fifth session write-up

The session included presentations and contributions from Milan Bollegala and Brendan Johnston on their PDS Handbook and maturity assessment for police forces, as well as a presentation from James West on Agentic AI.

PDS presentation – Handbook and maturity assessment discussion

Brendan and Milan provided an update on the ongoing development of the interoperability maturity assessment for policing. This work includes the creation of a detailed definition paper, a practical handbook and plans for a centralised repository of interoperability assets, including guidance papers, data standards, and APIs.

Milan outlined the structure of the maturity model across four levels: foundational, structural, semantic and organisational.

  • Foundational, focusing on secure connectivity

  • Structural, covering data standards and APIs

  • Semantic, relating to shared vocabularies

  • Organisational, encompassing agreements and policies

The aim is to provide comprehensive guidance and an engagement strategy for police forces to self-assess or seek PDS support.

The group discussion then covered the importance of encouraging suppliers to publish APIs and align with standards, as well as the need for strategic alignment with national initiatives. The group has also mentioned the need of review and publish relevant standards in collaboration with the National Standards Assurance Board.

Milan also mentioned that pilot with a police force is planned to test the maturity assessment. PDS will circulate session outputs and the handbook draft once available.

The group discussed the persistent challenges in mandating interoperability standards across policing, mentioning the limited application of the MAIT standard, the diversity of the sector, and the lack of enforcement. The experience with the MAIT standard showed that standards need to keep evolving and be widely adopted, or they risk becoming outdated.

The group acknowledged that technology vendors often set the pace for interoperability. They recommended using existing tools like APIs to make quick improvements now while continuing to work on longer-term standards. They also highlighted the need for strong, tailored communication to win stakeholder support, focusing on real benefits such as greater efficiency, improved safety, and cost savings. The successful rollout of Microsoft 365 was mentioned as an example of how good communication can drive change effectively.

Humberside Police’s integration platform was shared as a practical example of how interoperability can improve the experience for victims. By connecting multiple systems, the platform increased efficiency and was also used in areas like witness care. The group mentioned the importance of tracking these efficiency and effectiveness gains, and Milan planned to contact Scot Dunn in Humberside to get insights to support wider adoption. They also agreed that different messages should be tailored for different stakeholders, such as CIOs, CFOs, and Chief Constables, focusing on outcomes that matter to each, including cost savings, better policing results, or improved support for victims.

 

Agentic AI presentation

James West, chair of the AI sub-group within the JES committee, gave an overview of agentic AI. He explained that, unlike traditional AI agents, agentic AI can operate with more autonomy, take initiative, and plan across multiple steps.

You can download the slides below 



Agentic AI.pptx

The technical requirements discussed included large language models, memory systems, planning modules, tool integration, and feedback loops, with a strong emphasis on human oversight. The group highlighted the need for open APIs and data standards to ensure evidence can be trusted and reproduced, along with clear governance and safety measures. AI could be used for things like real-time intelligence gathering and automating processes, but participants noted current challenges, such as slow integration and the need to choose tools carefully. Risks were also raised, including criminal misuse of AI (fraud, deep fakes, romance scams) and the potential for autonomous agents to overwhelm systems.

The discussion also explored how AI and interoperability interact. Participants agreed that AI projects can drive improvements in interoperability, and vice versa. Using AI effectively in policing relies on having high-quality, compatible data, which makes improving interoperability a key priority.

James West described how AI is currently being used in policing, including predictive analytics, computer vision, digital forensics, and robotic process automation. Most applications are still focused on automation or managing knowledge, with more advanced agentic AI only just beginning to be explored. The group also discussed funding and pilot opportunities, such as the MPCC’s Test and STAR funds, as well as the challenges of moving from a proof of concept to full implementation.

The meeting wrapped up with a list of agreed follow-up actions. Milan will share session summaries, guidance on using the Knowledge Hub, and the draft interoperability handbook. Other actions include working with a police force to pilot the maturity assessment, arranging a meeting to discuss the assessment framework, locating and sharing the “what good interoperability looks like” paper, contacting Scot Dunn in Humberside for metrics, and providing guidance on Knowledge Hub access. Future sessions will revisit data and standards work, give updates on outputs like the handbook and maturity levels, and include reflective reviews to track progress and identify gaps.

To read the previous forum sessions, visit our Interoperability Hub.