Technology’s Effect on Advocacy
Fourth, challenge the admissibility of AI-generated reports where appropriate. Several district attorneys’ offices have already expressed concern. The King County, Washington, prosecutor’s office has instructed police not to use AI to write reports, stating: “We do not fear advances in technology — but we do have legitimate concerns about some of the products on the market now.” If prosecutors in your jurisdiction are accepting AI-generated reports without scrutiny, raise the issue. Fifth, file public records requests. Under the California Public Records Act, you can request a department’s Draft One usage policies, training materials, audit logs, and any communications with AI vendors about the product’s capabilities and limitations. This evidence supports both your individual case and any pattern-and-practice theory you may be developing.
what outputs they produced, and whether those outputs were independently verified by a human being.
Conclusion.
Police reports have always required scrutiny. We know that officers omit facts, shade language, and write narratives that favor the law enforcement perspective. But at least we knew who wrote the report. We could cross-examine the author. We could test their memory, bias, and credibility against the physical evidence, the body-worn camera footage, and the testimony of witnesses. Artificial intelligence changes that equation. When a report is co-authored by a machine that processes only audio, cannot see what happened, and is designed to produce polished prose, the risk is not just inaccuracy. The risk is that the report will sound more authoritative, more consistent, and more credible than the facts actually support, and that the officer who signed it will not be able to tell you which parts they wrote and which parts the AI wrote. Senate Bill 524 is a necessary first step. But it is only a first step. The burden is on us, as plaintiffs’ attorneys, to demand the records, preserve the evidence, and ask the hard questions. When accountability depends on the facts, we need to know who — or what — helped write the story.
Beyond Police Reports.
AI-generated police reports are only the beginning. Law enforcement agencies are already deploying artificial intelligence across a range of activities, including predictive policing algorithms that direct patrol resources based on historical crime data, facial recognition software used to identify suspects, automated license plate readers that track vehicle movements, and AI-powered analysis of social media activity. Each of these tools raises its own set of civil rights concerns, from racial bias embedded in training data to the chilling effect on First Amendment activity. For plaintiffs’ attorneys, the lesson is the same across all of these technologies. When law enforcement uses AI, demand transparency. Ask what systems were used, what data was fed into them,
32
Consumer Attorneys of California
FORUM September/October 2026
Made with FlippingBook - Online catalogs