AI Police Reports: Year in Review (www.eff.org)

In 2024, concerns over the deployment of AI in police report generation have intensified, particularly regarding Axon’s Draft One, a leading tool in the space. Despite its widespread use, many view AI-written police reports as unproven and lacking transparency, raising fears about accountability in the criminal justice system. In a notable development, the King County prosecuting attorney's office has prohibited the use of AI-generated reports, citing insufficient reliability and transparency. Axon's design choice to erase initial drafts of reports makes it nearly impossible to discern AI contributions, complicating efforts to ensure the accuracy of police narratives. Amid these challenges, there is positive momentum for regulatory efforts aimed at curbing the unchecked use of AI in policing. Recent legislation in Utah and California stipulates that police reports generated with AI must contain disclaimers, require accuracy certifications from officers, and retain initial drafts for transparency. These developments mark significant strides toward accountability, as they empower the public to scrutinize AI usage in law enforcement. As the conversation around AI in policing evolves, further legislative actions across more states are anticipated, reflecting growing concerns over AI's role in the justice system.