In a significant development for law enforcement, artificial intelligence (AI) is increasingly being adopted by police departments across the United States to streamline the process of drafting police reports. Officer Scott Brittingham of the Fort Collins, Colorado, police department, initially skeptical of new technologies, has found his report writing time dramatically reduced from 45 minutes to just 10 minutes by utilizing an AI-powered software called Draft One. This efficiency allows officers to dedicate more time to responding to calls for service and proactively preventing crime.
Draft One, developed by law enforcement technology company Axon, which also produces tasers and body cameras, is designed to create initial drafts of police reports. Axon reports that Draft One has become its fastest-growing product since its launch in 2024. The concept for Draft One emerged directly from the staffing shortages facing police departments nationwide. A 2024 survey by the International Association of Chiefs of Police revealed that agencies were operating, on average, at least 10% below their authorized staffing levels. Josh Isner, President of Axon, stated, “The biggest problem in public safety right now is hiring. You cannot hire enough police officers. Anything a police department can adopt to make them more efficient is kind of the name of the game right now.”
The rise of AI in police report generation is not without its critics. Legal experts and civil rights advocates have voiced significant concerns regarding potential biases, inaccuracies, and transparency issues inherent in AI-drafted reports. These concerns are particularly salient given the crucial role police reports play in the criminal justice process, influencing decisions made by prosecutors, defense attorneys, and judges. Andrew Guthrie Ferguson, an American University law professor specializing in technology and policing, emphasized that “Police reports are really an accountability mechanism. It’s a justification for state power, for police power.”
Draft One operates by leveraging transcripts from body camera footage to generate report drafts within seconds of an officer’s request. The software then prompts the officer to review and edit the draft, filling in bracketed blanks designed to ensure thorough review and correction of potential errors or the addition of missing information. Axon President Josh Isner affirmed, “It really does have to be the officer’s own report at the end of the day, and they have to sign off as to what happened.” The AI model utilized by Draft One is a modified version of OpenAI’s ChatGPT, which Axon has further trained to mitigate “hallucinations” – factual errors that AI systems can randomly generate. Axon also collaborates with a group of third-party academics, restorative justice advocates, and community leaders to garner feedback for responsible technology development and bias mitigation.
Beyond Fort Collins, police departments in Lafayette, Indiana; Tampa, Florida; and Campbell, California, have also adopted Draft One. In Fort Collins, Technology Sergeant Bob Younger initiated a pilot program with approximately 70 officers after a compelling demonstration of the tool. He noted, “I was blown away at the quality of the report, the accuracy of the report and how fast it happened. I thought to myself, ‘This is an opportunity that we cannot let go.’” The Fort Collins Police Department has since expanded access to all officers and estimates that Draft One has reduced report writing time by nearly 70%, effectively giving back valuable time to citizens.
Despite positive feedback from some prosecutors, the use of AI in police reports has encountered resistance. In September 2024, the prosecutor’s office in King County, Washington, decided against accepting AI-drafted police reports. In an email to police chiefs, they expressed concern that using the tool would “likely result in many of your officers approving Axon drafted narratives with unintentional errors in them.” Axon has responded by asserting its commitment to continuous collaboration with various stakeholders to guide the responsible evolution of Draft One and stated that its AI model is “calibrated … to minimize speculation or embellishments.”
The American Civil Liberties Union (ACLU) has also publicly recommended against the use of Draft One. Jay Stanley, a policy analyst with the ACLU Speech Privacy and Technology Project, stated, “When you see this brand new technology being inserted in some ways into the heart of the criminal justice system, which is already rife with injustice and bias and so forth, it’s definitely something that we sit bolt upright and take a close look at.” Concerns include the potential for errors in body camera footage transcripts due to accents or the absence of nonverbal cues impacting the accuracy of AI-generated reports. While Draft One includes automatic blank fields to prompt officer review, it is possible to submit a report without making any changes to the AI-generated draft. Furthermore, once a report is submitted, the original AI-generated draft is not saved, precluding the ability to review what an officer did or did not change, a practice Axon states mimics traditional, hand-written report processes.
A significant transparency issue revolves around whether defendants are informed that the police report in their case was drafted by AI. Draft One’s final reports include a customizable disclaimer by default, indicating AI assistance, but departments have the option to disable this feature. The Fort Collins Police Department, for instance, does not include disclaimers on its AI-generated reports. However, Ferguson advocates for “radical transparency” as the best practice, a sentiment echoed in Utah, where lawmakers passed a law earlier this year requiring police departments to include such disclaimers on final AI-drafted reports.
Ultimately, AI tools like Draft One are seen as aids rather than replacements for human officers. Officer Brittingham concluded, “My overall impression is that it’s a tool like anything else. It’s not the fix. It’s not replacing us writing reports. It’s just a tool to help us with writing reports.” As AI continues to integrate into various facets of society, its application in sensitive areas like law enforcement necessitates careful consideration of its benefits, limitations, and ethical implications to ensure accountability and maintain public trust.
