The AI landscape doesn't move in one direction — it lurches. Some techniques leap from experiment to table stakes in a single quarter; others stall against regulatory walls, technical ceilings, or organisational inertia that no amount of hype can dislodge. Knowing which is which is the hard part. The State of Play cuts through the noise with a rigorously maintained index of AI techniques across every major business domain — classified by maturity, evidenced by real-world adoption, and updated daily so you always know where you stand relative to the field. Stop guessing. Start knowing.
A daily newsletter distilling the past two weeks of movement in a domain or two — delivered to your inbox while the index updates in the background.
Each dot marks the weighted maturity of practices within a domain — hover for a brief summary, click for more detail
AI that analyses video feeds for people counting, crowd management, and anomaly or incident detection in public and commercial spaces. Includes crowd density estimation and behavioural anomaly alerting; distinct from workplace safety monitoring which targets occupational rather than general security contexts.
AI-powered video analytics for surveillance, crowd management, and incident detection remains a practice in transition: vendor maturity is accelerating, documented real-world ROI exists, but operational barriers persist and are now clearly articulated. The practice uses computer vision and deep learning to monitor feeds in real time—counting people, estimating crowd density, detecting behavioral anomalies, and alerting operators to incidents. Production deployments across 12+ industries (stadiums, hospitals, retail, memory care, data centers) have documented cost savings ($115K+ first-year in stadium deployments) and incident prevention (49-59% crime reduction, 67% crowd-incident reduction at major venues). Vendor ecosystems have matured: Hikvision DeepinView, Cisco Meraki Gen-3, and Axis Object Analytics now offer edge-embedded ML at scale with simultaneous multi-scenario detection (PPE, violence, gathering, intrusion).
However, two unresolved barriers now define the practice's ceiling. First, the false-alarm crisis: operator alert fatigue remains ubiquitous (83% of alerts in mature security operations are false positives; 74% of companies struggle to achieve value within 90 days of deployment). A critical evaluation gap was identified in 2026: frame-level VAD benchmarks show state-of-the-art AUC-ROC >52% but translate to event-level precision <10%, revealing fundamental misalignment between published metrics and operational surveillance. Second, post-deployment governance: while technical performance improves under controlled conditions, integration failures and operator disengagement compound in the 90-day window, driving organizational adoption failure even when algorithmic capability is present.
Regulatory pressure (EU AI Act, GDPR) and privacy governance gaps (always-on facial recognition, data hoarding, algorithmic opacity) further constrain public surveillance rollout. Deployment concentration remains high in occupancy, retail, and event safety; broader public surveillance adoption is stalled by convergence of these technical, operational, and regulatory barriers. This is a practice with proven tactical ROI in specific contexts, but without resolution of systemic false-alarm and governance barriers, it remains confined to experimental/early-adoption contexts rather than mainstream deployment.
Vendor momentum accelerates: Hikvision's DeepinView and Guanlan AI claim 90% false-alarm reduction; Cisco Meraki Gen-3 MV cameras integrate cloud dashboard with cross-camera tracking, license plate recognition, and query-driven investigation (April 2026); Axis Object Analytics supports 16+ simultaneous algorithms. The Hikvision DeepinMind NVR (32-channel, 16 simultaneous models per engine) exemplifies production-ready ecosystem maturity, with PPE detection, violence detection, behavior analysis, and fire/smoke detection embedded at recorder tier.
Real-world deployments validate tactical ROI in specific use cases: IntelliSee documents 13 deployments across 12 industries (stadium, hospital, school, retail, memory care, data center) with $115K+ first-year cost savings and incident prevention on existing infrastructure. Vidisky reports 49-59% crime reduction with 277K+ hours analyzed. Crowd density prediction deployments at MetLife Stadium (67% incident reduction), Singapore Changi (proactive staffing), and Coachella achieve 92% accuracy with quantified safety outcomes. Case studies demonstrate competence in controlled contexts: occupancy monitoring, equipment protection, event safety, fall detection.
However, the false-alarm barrier is now crystallized as the core adoption constraint. A 2026 technical analysis identified modular pipeline architecture (contextual guardrails, zone-aware confidence, per-stage instrumentation) as the necessary fix; monolithic detection+classification pipelines fail in crowded, variable environments. Critically, a peer-reviewed evaluation study (April 2026) exposed frame-level vs. event-level misalignment: state-of-the-art models achieve AUC-ROC >52% but event-level precision <10%, revealing fundamental gap between benchmarks and operational surveillance. Post-deployment governance failures compound the technical barrier: 74% of companies struggle to achieve value within 90 days; alert fatigue (83% false positives in mature SOCs) drives operator disengagement and undermines continuous improvement. Integration failures (VMS routing, access control context, incident management system connection) and undefined post-deployment roles accelerate this decline. Regulatory compliance (EU AI Act high-risk designation, GDPR data-impact assessments, facial recognition restrictions) increasingly constrains deployment scope. Production rollout remains concentrated in occupancy, retail, and event safety; mainstream public surveillance adoption stalled by unresolved false-alarm crisis, post-deployment governance, and regulatory barriers.
— Technical analysis documenting 74% post-deployment failure rate driven by false alarms and alert fatigue; modular pipeline architecture with contextual guardrails reduced false positives by order of magnitude in production systems.
— Meraki cloud integration enables AI-powered video search with license plate recognition, cross-camera person tracking, and investigation export; shifts operational workflows from playback to query-driven incident response.
— Deployment metrics: 49-59% crime reduction, 13M video tours analyzed, 277K+ hours footage processed, 3100+ incidents addressed; demonstrates human-in-the-loop proactive detection model at scale.
— Production-ready edge AI recorder with 16 simultaneous algorithm models per engine covering PPE, people counting, fire/smoke, falling, gathering, conflict detection; indicates ecosystem maturity with embedded analytics at recorder tier.
— Three major deployments (MetLife Stadium 67% incident reduction, Singapore Changi proactive staffing, Coachella stage safety) demonstrate crowd density AI at scale with 92% accuracy and validated safety outcomes.
— Critical assessment documenting false alarm crisis (up to 90% invalid alerts), facial recognition bias, EU AI Act enforcement, and industry shift toward trustworthy AI in surveillance systems.
— Axis released production-ready radar-video fusion camera with dual-level fusion mechanism reducing false alarms through integrated 60GHz radar module enabling detection beyond video limitations.
— Major deployment data: 1.6M remote monitoring events across 18,258 U.S. retail locations in 2025 showing 96.1% autonomous threat resolution and 95% false alarm filtering via AI-enabled video analysis.