Freedom vs. Safety: Axon CEO Rick Smith on the Future of Policing
Summary
- Axon’s breakthrough came after seven years of consumer-market failure, when it proved its Taser could stop an attacker rather than merely cause pain. The police pivot at the end of 1999 followed two failed launches and a final $1 million from family and an existing investor. Demonstrations made the sale: “We offered them $100 if they could stand up after being shot.”
- Regulatory and litigation exposure nearly destroyed the business even after explosive growth. Smith says market capitalization fell from roughly $2 billion to $100 million, while shares dropped from $33 to $7 in 2005. An SEC investigation halted police purchases; he says it ultimately ended with “no violations found,” but hundreds of lawsuits required an unusually aggressive defense.
- Body cameras and cloud software created Axon’s much larger platform opportunity. Friedberg cites a $35–40 billion valuation; Smith describes more than a million officers uploading two hours of footage each day. The transformation almost failed operationally before board member Hadi Partovi helped build the software organization: “a software business from scratch inside a weapons company.”
- Surveillance acceptance remains a constraint on the expansion of public-safety technology. Friedberg argues that concentrated state power troubles citizens even when individual searches have safeguards. Smith favors legitimate-case verification, penalties for misuse and homeowner control over shared footage: “These are my cameras. I control how they are used.”
- Axon’s robotics strategy combines a substantial ownership position with exclusive distribution rights. Smith says Axon bought almost half of Skydio and secured exclusive access in markets important to Axon, while preserving its independence and ability to raise capital. He believes “millions of drones” will be deployed in the US over the next few years.
- The disputed Taser-drone project continues, with human control presented as its central boundary. Following the advisory-board resignations Friedberg dates to June 2022, Smith still argues that remotely applied non-lethal force could improve on sending officers into gunfire. He proposes layered authorization and administrator approval—the least “user-friendly” technology in policing.
- Smith sees policing as essential infrastructure through a potentially disruptive AI transition. He predicts defunding would be “catastrophic,” while remaining optimistic about an “almost utopian society” in 10–20 years. That optimism comes with an explicit transition risk: technology may produce abundance before society comfortably absorbs the changes.
Deep dive
1. The consumer failure financed a police breakthrough
Smith traces the company to a Tucson garage in 1993, when he was 23 and the original Taser inventor was 73. Conversations about American gun violence—and remembering five or six people he knew who had been shot—prompted his question: “Why do bullets still exist in 1993, let alone 2026?” Captain Kirk’s phaser supplied the aspiration: stop a threatening person without killing them.
The original commercial thesis was consumer self-defense. Replacing gunpowder with compressed air could keep the new device outside firearm classification, allowing customers such as Smith’s mother to buy it. But consumers asked the devastating validation question: “If this thing really works, why aren’t the police using it?” After seven years and two failed launches, the company pivoted to police at the end of 1999.
When Smith proposed shutting down in 1998, his father explained that roughly $2 million owed to Silicon Valley Bank put his personal assets at risk. His remaining $500,000, matched by investor Bruce Culver, became the last $1 million. It funded the police product and survival through break-even. The 2001 IPO valued the company at about $12 million; Smith’s overriding relief was knowing his father would not lose everything.
The technical breakthrough was sufficient neural stimulation to immobilize someone, rather than pain a determined attacker could endure. Earlier generations had failed to stop focused, aggressive attackers; Smith says the Los Angeles police had tried a Taser on Rodney King, then switched to batons, and the device’s poor reputation was deserved. Axon used animal testing to adjust the stimulation. A former Marine toured police stations in a Winnebago, offering $100 to anyone who could remain standing after a demonstration. Officers immediately recognized situations where it could have prevented a shooting or violent struggle. Smith says the business reached break-even within a year.
2. Legal exposure overwhelmed the early growth story
Smith corrects Friedberg’s suggestion of immediate post-IPO stagnation: the business initially tripled annually, and he says its stock ranked first worldwide in 2004. An SEC investigation into product-safety representations then changed everything. Smith describes a short-seller attack involving leaked internal documents; he says his personal assistant was a mole who sent documents to the short sellers, who passed them to The New York Times. Shares fell from $33 to $7 in 2005, and police customers stopped purchasing because they could not defend buying from a company under federal investigation.
Market capitalization fell from approximately $2 billion to $100 million while lawsuits multiplied into the hundreds. “We should have gone bankrupt.” Smith brought boxes of documents to the SEC against legal advice, hoping to accelerate an investigation he says eventually ended with “no violations found.” His point is that surviving the inquiry did not reverse the commercial damage or litigation burden.
Smith rejected an insurer’s proposed roughly $2 million settlement, fearing it would attract more claims. Twice, Axon instead negotiated payments to itself and fought the cases, which he says it won. He reports zero lawsuits for the first time in a decade. The boundary remains conditional: fix genuine wrongdoing, but defend cases the company believes are unjustified “like religious fanatics.”
3. Body cameras required a second company inside the first
The camera idea grew from litigation. Smith watched officers struggle to convey a violent encounter while a deceased man’s family showed graduation photographs. Professional testimony could not reproduce what officers experienced: “We need to record this so you can show people what you saw and what you did, instead of trying to explain.” Police nevertheless resisted the “Big Brother” implications for another seven years.
Friedberg’s pushback is liability: recording can expose an officer’s misconduct, too. Smith answers with a Phoenix officer’s account of an arrested woman threatening a false rape allegation, then changing her behavior when told the interaction was recorded. His broader adoption lesson comes from a union leader: “No cop wants to wear a camera until he has worn it for 90 days,” after which he will not patrol without one.
Software nearly derailed the expansion. Smith credits college friend and director Hadi Partovi, effectively his part-time CTO for five years, with rebuilding the team and moving from an internal data center—where Axon nearly lost all customer data—to AWS and then Azure. Partovi also pushed the rebrand: “Taser is kind of a sledgehammer brand.” Shareholders had questioned turning a weapons company into a software company.
Smith identifies Ferguson as the catalyst: Michael Brown’s family and the police union both wanted cameras to resolve incompatible accounts. Smith says further investigation showed Officer Wilson was mostly correct, but it was impossible to prove what happened. Cloud infrastructure let Axon scale without asking 18,000 departments to secure their own servers. Smith describes more than a million officers uploading two hours each daily; Friedberg cites a $35–40 billion market capitalization. Smith argues footage often vindicates officers, a conclusion he says some former civil-liberties supporters resist.
4. Surveillance safeguards face a deeper objection about power
Friedberg raises municipal resistance to Flock cameras and the backlash to Ring’s Super Bowl advertising. His challenge extends beyond unauthorized searches: people may reject the existence of such comprehensive monitoring because the government’s power itself feels “creepy.” Explaining a useful application does not necessarily resolve that emotional or political objection.
Smith accepts the concern but argues society cannot return to the camera environment of the 1950s: cameras already outnumber people. The actionable question is data governance. Finding a stolen car or plates near a violent crime differs from tracking every American or investigating an officer’s former girlfriend. He describes Axon’s Ring partnership as enabling homeowners to share relevant footage while preventing police from casually browsing recordings.
Smith supports Senator Hawley’s bill, describing criminal and civil penalties for personal misuse and manufacturer checks that a search relates to a legitimate law-enforcement case. His example requires another officer to participate, increasing the personal stakes of an abusive search. He thinks effective safeguards can persuade 90% of people approaching the issue rationally, while acknowledging that agreement will never reach 100%.
5. Skydio extends Axon into the response itself
Smith says Axon has purchased almost half of Skydio, becoming its largest investor, and obtained exclusive channel rights in markets important to Axon. Keeping Skydio independent preserves startup speed and freedom to raise and spend capital without a public company’s constraints. The scale assumption is explicit: “We believe that millions of drones will be deployed in the US in the next few years.”
The operating proposition is earlier situational awareness. Drones can launch from stations around a city and sometimes arrive before a 911 operator finishes handling the call, reducing unnecessary dispatches and informing responding officers. Smith says agencies can publish mission profiles, and most prohibit random drone patrols: flights respond to specific requests for help. He also recounts people surrendering to drones before officers arrive.
6. The Taser-drone dispute exposes the limits of agreement
Friedberg revisits the advisory-board resignations he dates to June 2022, following plans for a Taser-equipped drone intended to address school shootings. Smith confirms research continues. His argument starts with the current alternative: sending armed people into a shootout. A robot can approach an armed attacker with non-lethal equipment without asking an officer to accept that personal risk. “Is this really the best we can do as a society?”
Smith thinks some advisers feared political backlash, and says several privately agreed with his reasoning. Axon reorganized the body into an Equality and Ethics Council with greater Black representation and continued participation from police critics. His design lesson is that affected communities need representation early: criticism can change safeguards before launch, whereas afterward “all we can do is argue.”
For robotic systems, Smith draws a categorical line: “Any decisions about using force against another person will be made by humans.” He describes layered authorization, configurable to require real-time administrator approval before even non-lethal force, and says Axon will insist on independent civilian oversight and verification every time a system is used. Unlike handing an officer a gun, the technology could involve several decision-makers before force occurs.
Friedberg’s unresolved challenge is concentrated state power and its potential use against citizens’ speech and rights. Smith counters with a mistaken-address raid: frightened armed officers now make split-second decisions; remote operators could work outside danger, record everything and use non-lethal tools. He acknowledges possible misuse, but rejects police arriving in a “’74 Chevrolet” while everyone else gains robotics and AI.
7. Smith expects public safety to anchor an unstable transition
Smith predicts actual police defunding would be “catastrophic.” He cites Montreal’s 1968 police strike over pay—not a George Floyd-like incident—and its “Night of Terror” as an illustration of how quickly order can unravel. He also invokes Steven Pinker’s account of the strike as evidence that a functioning police force is “the glue that holds society together.”
Smith says shoplifting decriminalization after the George Floyd events led to serious consequences, including abandoned parts of downtown San Francisco, disappearing jobs and reduced economic activity; he says he saw the decline himself. Friedberg responds that San Francisco has started reversing course. Smith says he tries not to get too involved in politics because corrective mechanisms can work over time, and argues that public safety enables other social contracts.
Earlier, Friedberg acknowledges that his own irritation at receiving three speeding tickets reflects a different daily experience from living amid violent crime, calling such views “luxury beliefs.”
AI makes the transition harder to forecast. Smith contrasts recent narratives of a jobs apocalypse with subsequent claims of millions of new jobs, concluding that people do not know what to expect. He thinks disease treatment and Elon’s idea of “unconditional high income” could help create an almost utopian society in 10–20 years, but explicitly anticipates possible difficulties getting there. Axon’s role, he says, is equipping police with robust controls.