Met Police Issues £72M Procurement for Digital Public Contact Platform Services

The Metropolitan Police Service, acting for the National Police Chiefs' Council, has issued a GBP72 million multi-year procurement for a single supplier to manage and operate England and Wales's Digital Public Contact platform from July 2027 to December 2029, with a possible three-year extension, The Register reported in May 2026. The contract covers ongoing service delivery for the Single Online Home websites, Police.uk, Data.police.uk, iHub, and the National My Police Portal; the Digital Public Contact Programme told Let's Data Science the GBP72 million figure is for service management and consolidation, not a standalone AI investment. AI features as one component: West Yorkshire Police went live with Post Call Analysis in March 2026, an NPCC-backed system that transcribes and categorizes archived control-room calls and identified 21% more calls with hidden-vulnerability indicators than prior manual processes during its pilot, per an NPCC statement. Humberside Police is separately piloting a natural-language call-handling switchboard.
The headline number, GBP72 million, is mostly a service-management and consolidation contract for platforms police forces already run; the more concrete AI story here is a named, already-live system with a measurable early result: West Yorkshire Police's Post Call Analysis pilot found 21% more calls contained hidden-vulnerability indicators than standard processes previously captured.
What happened
The Metropolitan Police Service, acting on behalf of the National Police Chiefs' Council (NPCC), has issued a procurement seeking a single supplier to manage and operate the existing Digital Public Contact platform under a GBP72 million multi-year service contract covering July 2027 to December 2029, with a possible three-year extension, The Register reported. The platform underpins the Single Online Home websites, Police.uk, Data.police.uk, iHub, and the National My Police Portal (NMPP), which is piloting GOV.UK One Login. After the story published, the Digital Public Contact Programme told Let's Data Science by email that the GBP72 million figure covers ongoing service management and consolidation of these existing platforms, not a standalone AI deployment, and that AI is one component within the broader service scope.
Technical context
The most concrete AI element is Post Call Analysis (PCA), which West Yorkshire Police and the NPCC's Digital Public Contact programme took live on March 17, 2026, making West Yorkshire the first force in England and Wales to deploy it, per an NPCC press release. PCA transcribes and categorizes historic control-room calls inside a police-controlled environment, flags repeat-caller patterns, detects indicators of hidden vulnerability that operator notes may miss, and checks compliance with crime-prevention and forensic-preservation advice. The NPCC says PCA is trained exclusively on police data and pairs outputs with plain-language explanations in line with national AI-in-policing standards; during pilot testing it identified 21% more calls with hidden-vulnerability indicators than standard processes previously captured. Separately, Humberside Police is piloting a natural-language switchboard, and Cambridgeshire chief constable Simon Megicks described a live-assist tool that "listens in and conducts live database searches, surfacing relevant information instantly" during calls, per The Register.
For practitioners
Multi-year service-consolidation contracts of this size matter less for their headline value than for what they standardize: a single supplier managing several existing public-contact platforms creates one integration and security-review surface rather than several. Within that, PCA is a useful reference design for regulated, high-stakes transcription and categorization work: narrow scope limited to call analysis rather than decision-making, training data limited to in-domain police data, plain-language output explanations, and a published, measurable pilot metric rather than only qualitative claims.
What to watch
- •Whether the GBP72 million procurement's winning supplier and contract documents disclose privacy, retention, and audit-logging requirements for the AI components it will manage.
- •Wider rollout of Post Call Analysis beyond West Yorkshire Police, and whether the 21% vulnerability-detection improvement holds at scale.
- •Outcomes from the Humberside Police natural-language switchboard pilot and the live-assist call-handler tool referenced by Cambridgeshire's chief constable.
Editorial analysis
The Digital Public Contact Programme's clarification, that this is a service-consolidation contract with AI as one component rather than a national AI deployment programme, is itself a useful corrective to headline-driven framing of public-sector AI procurement: the largest line item in a contract is not always the most consequential capability within it, and PCA's narrow, measured pilot, a specific, published percentage improvement, is a more concrete signal of AI impact here than the GBP72 million topline.
Key Points
- 1West Yorkshire Police's Post Call Analysis pilot identified 21% more calls with hidden-vulnerability indicators than prior manual processes, per NPCC.
- 2The Digital Public Contact Programme clarified that the GBP72 million procurement is a service-consolidation contract, not a standalone AI investment.
- 3Humberside Police and a Cambridgeshire-referenced live-assist tool show AI call-handling support is spreading beyond West Yorkshire's initial pilot.
Scoring Rationale
A GBP72 million national procurement plus a live, measured AI pilot (Post Call Analysis, a 21% improvement in detecting hidden-vulnerability indicators) is a concrete, well-documented public-sector AI adoption story relevant to practitioners in regulated, high-stakes deployments. The Digital Public Contact Programme's own clarification that AI is one component within a broader service-consolidation contract, not the primary spend, moderates the story's scale without reducing its operational relevance.
Sources
Public references used for this report.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Ad Tech problems