Axon brings real-time AI alerts to its next body-camera platform

What Axon announced

Axon announced two new body-camera models on October 24: Body 5 and Body 5 Pro. The company says both add substantially more on-device computing than its earlier cameras, along with improved low-light video, location reporting, battery endurance and network support. The Pro model also adds a short-range camera intended to scan a physical driver’s license and return license or warrant information on the device.

The more consequential change is Guardian, a set of real-time alert features tied to Axon’s AI Era plan. Axon says its first body-camera feature can analyze audio while a camera is recording, identify signs that an encounter may be escalating and alert a connected operations center or nearby officers through Axon Fusus.

Human review and availability limits

According to Axon, Guardian creates an alert rather than choosing a response. A person reviews the livestream and decides whether action or additional support is appropriate. That distinction is central in a public-safety setting, but the announcement does not provide accuracy measures, false-alert rates or independent field-test results.

Axon has not announced general availability dates for Body 5 or Guardian. The reviewed pages also do not publish prices or a country-by-country rollout. Features vary by configuration, and Guardian is presented as part of the separate AI Era plan. Agencies therefore cannot treat the announcement as confirmation that a specific configuration is available for purchase or deployment.

CGH perspective

Real-time analysis moves body-camera AI from after-the-event paperwork toward decisions during an encounter. Any evaluation should test alert accuracy across noise, accents and stressful speech, document when analysis begins and ends, and define who can view a livestream. Procurement teams should also confirm retention rules, network dependencies and escalation procedures. Human review is a meaningful control only when reviewers receive enough context and agencies measure both missed events and unnecessary alerts.