From the Public Safety edition of October 21, 2026
NPCC innovation awards report West Yorkshire Police’s AI post-call analysis at 97.6% accuracy versus 82% for human reviewers, and Cleveland Police’s AI redaction pilot cutting processing time 85–90%
National Police Chiefs’ Council and Police Digital Service; West Yorkshire Police; Cleveland Police · Policing / control rooms and digital evidence · England and Wales, United Kingdom
- Publisher
- National Police Chiefs’ Council
- Original publication
- October 21, 2026 (NPCC news release)
- Source retrieved
- 2026-10-22
What happened
The NPCC release on its Innovation and Digital Summit awards says Post-Call Analysis, built by West Yorkshire and Hertfordshire police, the NPCC Digital Public Contact Programme and Deloitte, turns control-room audio into transcripts and summaries and flags vulnerability. Evaluation found 97.6% accuracy on key call characteristics versus 82% for human reviewers, vulnerability indicators identified in 99.3% of relevant cases, about 5.7 times more context captured and about 150 times more calls processed per hour; West Yorkshire deployed it operationally. Cleveland Police’s Pimloc Redact pilot cut per-file processing by about 85–90%, with human quality review. Warwickshire reported 42% of staff using Copilot Chat within eight weeks.
Why it matters
These are among the more specific measured results for police AI in contact centres and evidence handling, though reported in an awards release rather than a published evaluation. They give U.S. dispatch-adjacent police call review and records units concrete metrics to ask vendors for, and show human review retained in both deployments.
Evidence and measured results
Opened the NPCC release dated Oct. 21 and read all award entries, including the PCA accuracy and throughput figures, the Cleveland redaction pilot metrics and user ratings, Thames Valley/Hampshire automation savings and Warwickshire Copilot adoption figures.
Limitations and uncertainty
Awards release, not the evaluation reports; methods, sample sizes and baselines for the PCA and redaction figures are not given; accuracy ratings for redaction are user-reported.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-10-22; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Control rooms and evidence units cannot manually review or redact the growing volume of recorded audio and video.
- Stakeholders
- Control room managers
- Digital evidence or disclosure units
- Professional standards
- Legal services
- Chief information officers
- Discovery questions
- What share of calls is reviewed today?
- How long does a typical redaction take?
- Who checks AI output before release?
- Is there a baseline to compare against?
- Value hypothesis
- Human-reviewed AI analysis and redaction can raise review coverage and cut handling time, if local tests confirm accuracy.
- Potential engagement
- Eight-week measured pilot of call analysis or redaction against a manual baseline.
- Unsupported claims
- That 97.6% accuracy will transfer locally
- That redaction accuracy is independently validated
- Specific cost savings
Pre-sales engineering
Role takeaway
- Fit
- Agencies with recorded call audio and large video or audio evidence volumes.
- Architecture and integration
- Connect call-recording and evidence-management stores to transcription, classification and redaction services with write-back.
- Prerequisites
- Access to recordings
- Human-labelled test sample
- Records system integration path
- Constraints
- Recording formats and retention vary
- UK results may not transfer to U.S. systems
- Disclosure rules differ
- Security
- Keep sensitive audio and evidence in approved hosting with access logging.
- Proof of value
- Proposed: compare AI classification with a labelled call sample
- Proposed: time redaction before and after with reviewer error checks
Delivery
Role takeaway
- Work
- Baseline current effort, run the pilot with reviewer checks, and report accuracy and time results.
- Dependencies
- Recording access
- Reviewer availability
- Vendor support
- Ownership
- Control room manager or digital evidence unit lead
- Skills and adoption
- Train reviewers on verifying AI flags and redactions.
- Governance checkpoints
- Accuracy threshold agreed before pilot
- Error review before wider rollout
- Acceptance criteria
- Proposed: classification accuracy at or above manual baseline on labelled sample
- Proposed: redaction time reduced with no missed redactions in audit
- Risks
- Missed vulnerability flags
- Over-reliance on automated redaction
- Unpublished evaluation methods
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Post-call analysis requires access to recorded control-room audio, transcription, classification and a searchable store; redaction tools work on video, audio and digital evidence within evidence-management workflows.
Governance
Who approves, reviews and stays accountable for outcomes?
The redaction pilot retained user review and quality assurance; the release does not describe governance for PCA vulnerability flags or how they affect call handling.
Security and privacy
What data, permissions and controls need testing?
Both tools process sensitive call audio and evidence; the release does not describe hosting or data protection arrangements.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Redaction users reported saving two to four hours per job; Warwickshire paired Copilot rollout with training and peer sessions.
Procurement
What should contracts, pricing and exit terms secure?
PCA was developed with Deloitte; Cleveland piloted a commercial product (Pimloc Redact).
Operating model
Which teams own the service once it runs?
PCA shifts call review from sampled manual checks to automated analysis of far more calls.
Publication history
- 2026-10-21Public Safety · Issue 438 resources
Stable resource ID: npcc-innovation-awards-west-yorkshire-post-call-ai-cleveland-redaction-20261021