Axon Orders AI Use, Ties Stock Grants to Productivity Targets

Axon Enterprise expects every product management, design, and software engineering task to be compressed by 25% to 30% by the end of 2026. Jeff Kunins, the company's chief product officer and chief technology officer, told Fortune the change will require less labor and organizational capacity than building the same projects without AI.

The mandate lands as the Taser maker's AI surveillance tools draw mounting criticism. Axon has reported ten consecutive quarters of revenue growth above 30%, driven partly by software and artificial intelligence sales. But the same product line has put the company at the center of a national debate over police technology.

How the mandate works

Kunins described a carrot-and-stick scheme aimed at Axon's research and development team of roughly 2,200 employees:

• Every team must hit a 10% mandatory target in productivity gains. That baseline means teams get fewer resources to advance new features and products.
• Teams that prove gains above 15% receive stock grants, provided the broader team clears the 10% minimum threshold.

"We're expecting to yield about 25%-ish acceleration, not just of the coding part, but of the entire person effort going into these features," Kunins said.

Internally, Kunins runs hackathon training sessions that pair engineers with the wider employee base. The IT department customizes AI features from vendors including Workday and Salesforce to automate workflows.

Why the timing is fraught

The push into AI coincides with escalating scrutiny of how officers use these systems. Police officers have been fired or charged with using Flock Safety's AI-enabled cameras to stalk former romantic partners. AI systems erroneously led to the 2023 arrest of a man accused of trying to lure a child from a McDonald's, and more recently sent a Florida woman to jail for two weeks after she was wrongly accused in a deadly hit-and-run.

Most public anger has centered on Flock. Axon has faced its own criticism: Forbes reported that the company's AI police-report tools make errors, though Axon said those problems occurred during an early access period when agencies could provide feedback before the broader launch. Earlier this month, the Los Angeles City Council voted to delay an Axon pilot program that would have attached AI-enabled license plate readers to streetlights.

In August, Axon founder and CEO Rick Smith acknowledged the budding anti-surveillance movement during the company's second-quarter earnings presentation. He compared it to the backlash that followed the release of Tasers.

"Any new technology in public safety is going to go through an adjustment period, and I would say the critics aren't all wrong," Smith told investors.

Kunins said the public discourse around surveillance and AI demands nuance.

"I am personally a life-long, die-hard progressive, and I'm here because of that, not in spite of it," Kunins said. "We view our work as using technology to improve transparency and accountability on both sides…and our role is to build the tech that facilitates public safety."

What happens when the AI gets it wrong?

Axon runs new features past a third-party group called the Ethics and Equity Advisory Council, which makes recommendations during product design and development. The process surfaced a real problem in Draft One, the generative AI tool that converts body camera audio into a first draft of a police report.

Kunins said the large language models produced text saying a white subject "left the scene," while a Black person "fled the scene," even when all other factors in the case were identical.

"Obviously, that's a flagrant example of inappropriate bias coming out of the models solely because of the description of the skin color of the person," Kunins said. Axon fine-tuned the models to make the outputs more consistent and remove harmful bias.

The company has also deployed AI in less contentious ways. Axon Assistant translates more than 50 languages and dialects captured by body cameras, letting officers communicate in the field without a human interpreter. For nearly a decade, Axon has used AI to redact camera footage, blurring bystanders' faces, license plates, and other sensitive details before release to lawyers, the press, and the public. Before that, an employee had to go through every frame by hand.

The bet behind the targets

Smith founded Taser International in an Arizona garage in 1993, redesigning a NASA scientist's device for law enforcement. The company, now called Axon, has sold more than one million Taser devices and expanded into body cameras, drones, and license plate reader cameras for police, emergency responders, and private security customers.

The AI mandate extends that playbook: cut the cost and time of building products while the products themselves remain under public scrutiny. If teams hit the 25% acceleration Kunins expects, Axon gets more software shipped with the same headcount. If the bias and error problems resurface at scale, the company will face the same adjustment period Smith predicted in August.