The State of AI Readiness in Physical Security: What 300 Security Leaders and About 460,000 Devices Show About Readiness for AI
Cameras now ship with AI that runs on the device itself — plain-language detection, on-device neural processors, licensed analytics apps — and using any of it depends on your team’s ability to patch, configure, and control every device.
This report combines a 300-leader survey with Platform Preview data covering about 460,000 devices to show how ready physical security teams actually are for the AI that’s already arriving on their fleet.
Key takeaways:
- Why the AI manufacturers are shipping lands on the same devices teams already struggle to see and update — 80% of devices carry known vulnerabilities, 71% are EOL or approaching end-of-service, and 100% of vulnerable supported devices had a patch available but still weren’t current
- How AI amplifies whatever it’s built on across three dimensions — data (confident answers from incomplete records), automation (one bad decision across every device at once), and governance (the audit trail, rollback path, and named owner CISA’s December 2025 OT guidance now calls for)
- The variable that tracks AI readiness most closely in the data: who owns device management. Teams managing devices mostly in-house run far ahead of teams relying on outside support on automation, visibility, and cross-team integration — and 73% of outsourced teams would prefer direct control
- Where the findings sit on the AI Readiness Roadmap — most teams are still in Phase 1 (Device Health & Compliance) — and the five action guardrails to put in place before the next automated change goes live: authorization, audit trail, named ownership, human in the loop, and active log review
This resource is for you if:
You lead physical security, IT, or cybersecurity at an enterprise where AI is already a 2026 project focus — or is about to be — and you need a clear-eyed read on where your device foundation stands before automated and AI-driven actions start reaching the fleet.