San Diego police department considers AI for writing reports

By Splitfeed EditorialLast updated Oct 14, 2026, 15:10 UTC

Between October 12 and 14, 2026, multiple outlets reported that the San Diego Police Department is actively evaluating artificial intelligence tools to help officers draft police reports. Coverage from the AP, the Los Angeles Times, the Washington Post, and local outlet 10News confirmed the evaluation is underway, with at least one report noting the potential use extends beyond documentation to officer training as well. The San Diego officers' union has formally raised concerns about the proposal, making labor opposition part of the public record from the outset.

The story matters because police reports are legal documents. They form the foundation of prosecutions, civil litigation, use-of-force reviews, and disciplinary proceedings. Any alteration to how those documents are produced touches questions of accuracy, accountability, and officer liability in ways that go well beyond a routine efficiency upgrade. San Diego is not the first department to explore this, but the union's early and public objection signals that the implementation path will not be straightforward. What remains genuinely in dispute is whether AI-drafted reports would reduce officer error and administrative burden or introduce new risks around factual distortion, bias in language models, and the assignment of legal responsibility when a report contains an error.

None of the three camps have published specific reactions to this story yet, so what follows reflects what each would typically argue. The Pro-AI camp would likely point to the hours officers spend on paperwork as time taken away from patrol and community work, framing AI assistance as a tool that lets officers do more of what they were trained for. The Anti-AI camp would be expected to raise concerns about algorithmic bias entering the official record and the difficulty of holding anyone accountable when a machine produces a document used in court. The Middle Ground camp would probably argue that AI assistance could be acceptable under strict conditions, such as officer review and sign-off requirements, clear audit trails, and negotiated union agreements that protect workers from arbitrary discipline based on AI-flagged inconsistencies.

The clearest signals to watch for are whether the San Diego city government or police department releases a formal evaluation timeline, what specific conditions the officers' union sets as requirements for any agreement, and whether the department names the vendor or tool under consideration, which would allow scrutiny of how similar systems have performed elsewhere.