Summary: Part 2 of our Verkada ISC West 2026 coverage takes us through their AI gallery, where six of their most impactful AI features were showcased in an interactive museum-style display. Key highlights include compound alert triggers for workplace safety violations, fighting detection, automatic slip and fall detection with AI summary, and their AI deterrence feature, which uses scene analysis and personalized talk-down messaging to deter unwanted activity before an alarm is ever triggered. With 174 AI updates released in the past year alone, Verkada’s platform is evolving rapidly and the AI gallery at ISC West gave a clear picture of where that development is heading.
If you caught Part 1 of our Verkada coverage, you already know that their ISC West 2026 booth was built around an interactive Louvre museum experience designed to put their AI-powered investigation tools through a real workflow. Part 2 picks up where that left off, moving from the crown jewel exhibit into Verkada’s AI gallery, where six of their most significant AI features were on display.
The gallery itself was a creative continuation of the museum theme, replacing traditional paintings with live demonstrations of Verkada’s latest AI use cases. Here is what was on the walls.
174 AI Updates in a Year
Before getting into specific features, that number is worth acknowledging. Verkada released 174 AI updates to their platform in the past year. The AI gallery at ISC West represented a curated selection of what they consider their greatest hits from that development cycle. For organizations evaluating a long-term security platform, that pace of development matters. The platform an organization deploys today is not the platform they will be running two years from now, and with Verkada, that trajectory is clearly moving in one direction.
Compound Alerts: Alerts That Require More Than One Condition to Trigger
One of the most practical AI features on display was compound alerting, and the workplace safety example used to demonstrate it was straightforward. The system was configured with two conditions: a person crossing a defined line and that same person not wearing a safety vest. Only when both conditions were true simultaneously did an alert get generated.
That logic eliminates the noise that makes single-condition alerting unreliable in real operational environments. A person crossing a line while properly equipped is not an issue. A person crossing that same line without the required safety gear is. Compound alerts let security and operations teams define the specific combination of circumstances that actually warrants attention, which directly reduces alert fatigue.
The applications extend well beyond workplace safety. Organizations can configure compound alerts to notify them when a specific person of interest enters a defined area, when a VIP customer arrives at a particular part of a facility, or when any set of conditions that matters to their operation occurs together. The logic is fully customizable.
Additional AI features on display in the gallery included:
- Fighting detection that identifies physical altercations in camera footage and triggers an alert without requiring an operator to be actively watching the feed.
- Automatic slip and fall detection with auto-archiving and an AI-generated summary of the incident, which has direct applications for liability documentation in retail, healthcare, and hospitality environments.
AI Deterrence: Proactive Rather Than Reactive
The feature that generated the most conversation at the Verkada booth was AI deterrence, and for good reason. It represents a meaningful shift in how security technology responds to threats, moving from a reactive posture to a proactive one.
The concept is built around a straightforward observation: security teams spend significant time monitoring high-risk areas so that they can respond verbally through a speaker if someone enters an area they should not be in. That process requires a live operator to be alert and watching at the exact right moment. AI deterrence automates that response entirely.
When the system detects an unauthorized presence, it performs scene analysis of the camera feed and generates a personalized, tailored talk-down message describing what the person is wearing, what they are holding, or other visible characteristics. Rather than a generic pre-recorded warning, the message is specific to the individual in the frame.
The workflow escalates through three successive messages if the person does not respond:
- A polite initial message, such as notifying the individual that the location is closed and directing them to return during business hours.
- A more direct second message that references the person’s appearance and explicitly states that authorities will be contacted if they do not leave.
- A final escalated message with stronger language and further personalization.
If the individual ignores all three warnings, the system can escalate all the way to triggering an alarm and initiating police dispatch, depending on how the customer has configured the workflow.
Three primary use cases have emerged for AI deterrence:
Protecting outdoor assets — Equipment yards, vehicle lots, construction sites, and any facility with valuable assets stored outside are particularly well suited for this feature. The deterrence message intercepts a potential theft before it occurs rather than documenting it after the fact.
Preventing crimes of opportunity — Facilities with valuable goods inside can use AI deterrence outside as an early warning layer that operates before the alarm system is ever triggered. The goal is to deter a break-in attempt before it reaches the point of an actual intrusion.
Managing loitering, vandalism, and illegal dumping — Not every unwanted behavior is a break-in attempt. Open areas that attract loitering, vandalism, or illegal dumping present a different kind of challenge. AI deterrence can address those lower-severity, higher-frequency situations automatically without requiring staff intervention.
The live demonstration at the booth gave attendees a firsthand look at how the personalization works in practice. The system identified clothing colors, described items being held, and delivered contextually relevant messages in real time. It was a clear illustration of how far scene analysis has come and what that capability makes possible in an operational security context.
What This Means for Your Security Strategy
What the Verkada AI gallery demonstrated is a platform that is actively closing the gap between detection and response. Compound alerts reduce the noise that leads to fatigue. AI deterrence addresses threats before they escalate. Automatic detection of incidents like slip and falls reduces the manual documentation burden on security and operations staff alike.
For organizations that are still primarily operating in a reactive security posture, these capabilities represent a concrete path to something more proactive and more manageable for the teams responsible for running it day to day.
Ready to explore how Verkada fits into your security program? Contact the NCS team to schedule a consultation and we can walk through the platform and what it would look like for your environment.
Learn more about Verkada and their full platform at verkada.com.
