Organizations can now configure their cameras to trigger an alarm when a vehicle crosses a digitally drawn line in the specified direction. This allows for faster response times as the alarm is raised as soon as a vehicle enters the site, rather than waiting for people to get out. Like other camera triggers, vehicle line-crossings are verified by a human agent before raising an alarm. Please refer to this article for a list of cameras that support vehicle line-crossing.

In addition to person detection, both line-crossing and loitering triggers are now supported on third-party cameras via Command Connector. This allows organizations to leverage their existing hardware to detect and deter unwanted activity such as trespassing and vagrancy while reducing false alarms from people passing through. To learn more, read the full announcement.

AI-powered deterrence is now supported on third-party cameras via Command Connector. With this capability, organizations can trigger a sequence of AI-generated warnings that can incorporate scene context and escalate over time if the person remains. Whether it's deterring illegal dumping in a parking lot or discouraging loitering outside a retail location, organizations can put their existing hardware to work as an active layer of perimeter protection. To learn more, read the full announcement.

Admins can now generate reports of alarm events across sites, with filters for time, location, device type, response level, whether dispatch was requested, and more. Reports can be easily exported as CSVs to streamline audits and record-keeping.

Admins can now mute specific sensors or cameras for a period of time from 30 minutes to eight hours. This can be used to prevent repeated false alarms from authorized activity such as overnight deliveries without disarming the entire partition or muting a device indefinitely.

Admins can now create arm/disarm schedules by simply describing them in plain language, such as “every weekday from 12-6am.” Schedule changes can also be made in natural language, such as “Fridays should disarm at 8:30 instead.”

Now in public beta, AI-powered deterrence enables organizations to stop unwanted activity with human-like warnings. Instead of a single robotic message, the system triggers a sequence of realistic AI-generated messages that can incorporate scene context and escalate over time if the person remains in the area. To learn more, read the full announcement.

Each deterrence event will now use a randomized voice to prevent repeat offenders from becoming accustomed to the same voice. This approach reinforces the perception of active monitoring and helps ensure that deterrence remains credible even in locations with frequent or recurring activity. To learn more, read the full announcement.

Organizations can now enable AI-powered deterrence on Verkada intercoms. For aesthetic-conscious locations such as retail stores, banks, and professional services firms, intercom-based deterrence offers a cleaner and more discreet installation.

Beyond performing AI deterrence, intercoms now act as full-fledged speakers in a Verkada alarm system – meaning they can be used for automated announcements and human agent talk down if an alarm is raised. To learn more, read the full announcement.

With two-way audio communication via the BK22, SOC teams now have a direct line with staff through the keypad’s built-in speaker and microphone. During an incident, security teams can initiate the call and speak directly with people on site to confirm safety, provide instructions, and quickly determine the appropriate response. To learn more, read the full announcement.