Axon president Josh Isner on AI-powered policing, the $630M Prepared acquisition, and taser drones for schools

Summary

Axon president Josh Isner used a wide-ranging conversation to lay out a company in the middle of a deliberate pivot from hardware manufacturer to AI-powered public safety platform, with acquisitions and product bets that signal where it sees the most durable revenue.

The Software Turn

For its first 25 years, Axon was primarily a hardware business — tasers, body cameras, physical gear. The last five years have shifted the center of gravity toward software. The company now manages more than 40 times the content volume of the entire Netflix library, all of it police evidence stored in the cloud. The flagship AI product to emerge from that data asset is Draft One, a generative AI tool that transcribes body camera audio and produces the first draft of a police report automatically. Isner puts current report-writing at roughly 50% of an officer's working time and believes Draft One can compress that to around 10%.

The Prepared Acquisition

Axon closed the acquisition of Prepared AI for $630 million, led by founder Michael Chime. The strategic logic is response-time compression. Today, the chain from a 911 call to an officer arriving on scene runs roughly two minutes, cycling through a call taker, a dispatcher, and officer routing. Axon's goal is to automate that entire sequence — using call metadata to launch a drone before a human dispatcher has made a decision, and ultimately routing officers with minimal human intervention. Prepared's 911 infrastructure is the entry point into that workflow.

Axon also cited a prior acquisition, Fusus, as part of the same platform-building approach, describing both deals as bets on teams with strong product synergy that can be accelerated by plugging into Axon's existing customer ecosystem.

Taser Drones for Schools

Isner's most forward-looking hardware bet is deploying less-lethal force on autonomous drones. The specific scenario he describes is school shootings: drones nested in school rooftops, equipped with taser technology, capable of flying autonomously to a threat and engaging it with targeting software before first responders arrive. He is explicit that a human, specifically a police department, would retain authority over when force is actually deployed. Axon frames this as closing the window between when a mass shooting begins and when it ends.

Facial Recognition: Interested but Not Yet In

The United States remains one of the few markets where facial recognition is not formally used in policing, and Axon has not yet entered that market. Isner says the company is now more interested than it has been, but the threshold condition is demonstrable removal of racial bias from the underlying algorithms. He expects the adoption debate to be resolved through a combination of major city police chiefs, sheriffs, mayors, and city councils rather than any single regulatory body.

Manufacturing and the Buy American Position

All Axon hardware, including tasers and body cameras, is manufactured in Scottsdale, Arizona. Isner notes some internal uncertainty about the company's long-term headquarters situation but is firm on domestic production. Taser cartridge manufacturing is now fully automated, with robots producing millions of units annually. He draws an explicit parallel to the process-knowledge premium seen in AI talent, arguing that engineers who run manufacturing automation well carry similarly concentrated, hard-to-replace value inside the company.

Market Realities

Isner acknowledges a structural constraint that keeps most startups out of this space: US public safety represents fewer than one million total users. That TAM is too small for most venture-backed companies to justify. Axon's position, built over 16 years, depends on capturing the majority of those users on a single platform and expanding revenue per user through a suite of products it calls capture to courtroom. The company operates an open API posture, which Isner frames not as a competitive concession but as a mission-driven necessity — a closed data environment in policing directly undermines crime-fighting effectiveness.