The Milwaukee Police Department (MPD) announced that it is actively evaluating generative artificial intelligence (AI) tools for use in police reporting. While the department is not yet deploying the technology, a new policy released in January outlines how AI‑generated drafts would be handled and stresses human oversight.
Potential benefits for officers
Prosecutor Jeffrey Altenburg, chief deputy district attorney for Milwaukee County, acknowledged that AI could be a helpful aid. “It can be a helpful tool as long as it’s overseen and regulated by actual humans making actual observations and making sure that what is generated is consistent with what was intended to be,” he said. MPD officials echo this sentiment, noting that AI could speed up the drafting of routine reports, allowing officers more time for community patrols and direct engagement with residents.
Safeguards built into the policy
MPD’s generative‑AI policy requires that any AI‑produced material be labeled as a draft and reviewed for accuracy before it becomes part of the official record. Officers must also indicate in the final work product that AI assistance was used. The department’s Generative Artificial Intelligence Committee has been meeting to assess a range of products, weighing utility, cost, accountability and impact on the quality of police documentation.
Civil‑rights concerns
Critics caution that large‑language models can produce “errors, hallucinations and bias,” especially when trained on data sets that reflect historical inequities. Amanda Merkwae, policy and advocacy director for the American Civil Liberties Union of Wisconsin, warned that AI‑generated reports could create credibility issues for officers and potentially place them on the Brady list, a record of law‑enforcement personnel whose testimony may be questioned in court.
Merkwae also raised a scenario in which an officer’s testimony conflicts with an AI‑assisted report, asking whether the discrepancy would be attributed to a technical glitch or to the officer’s honesty. “If it was a human‑written report and there’s that type of contradiction that can rise to the level of a potential dishonest statement — that could land a prosecutor putting a police officer on the Brady list,” she said.
Practical limits highlighted by prosecutors
Deputy District Attorney Matthew Torbenson stressed that AI cannot replace the nuanced observations officers make on scene. He cited a child‑neglect investigation where details such as the smell of a home, evidence of a bug infestation, or the availability of food must come directly from the officer’s senses. “All of that has to come from the impressions of the officers that are on scene,” Torbenson explained.
Balancing innovation with accountability
MPD’s spokesperson reiterated that the department is still in the “consideration phase.” “Any onboarding of an AI product is only in the consideration phase, so have we given it thought? Yes. Are we anywhere close to implementing that technology? No,” the spokesperson said.
Forensic science practice coordinator Connor Romenesko of the Wisconsin State Public Defenders Office warned that rushed or unchecked AI drafts could unintentionally lead to wrongful charges. He described a scenario where an officer, fatigued at the end of a shift, fails to thoroughly review an AI‑generated report, and a prosecutor then relies on that incomplete information.
Next steps
The department plans to continue its review, incorporating feedback from law‑enforcement leaders, prosecutors, civil‑rights advocates and technology experts. If the committee determines that the benefits outweigh the risks, MPD would move forward with a pilot program that includes strict oversight protocols and transparent reporting to the public.
Milwaukee residents can expect updates as the department refines its approach, ensuring that any AI tools adopted serve the community’s safety while respecting constitutional rights and the integrity of the criminal‑justice process.
Original reporting: Wisconsin Watch — read the source article.
