Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Police Audit topic
No spam. Unsubscribe anytime.
Independent audit: majority of Las Cruces Police investigations handled appropriately, reviewers flag timeline delays and AI report‑writing risks
Summary
The OIR Group presented its eighth semiannual audit to the council, reporting that the Las Cruces Police Department’s internal investigations were generally thorough but highlighting delays in administrative timelines for two critical incidents and noting emerging concerns about officers using AI to draft reports.
Get email alerts on the Police Audit topic
No spam. Unsubscribe anytime.
An independent police‑practices reviewer told the Las Cruces City Council work session on Monday that the Las Cruces Police Department’s internal investigations reviewed in the first half of the year were generally complete and appropriate, but that the department should shorten administrative timelines for critical incidents and formalize controls for emerging artificial‑intelligence tools used to draft reports.
The OIR Group, an independent outside reviewer, presented its eighth semiannual report covering internal investigations closed between Jan. 1 and mid‑year. Stephen Connolly of OIR Group said the firm reviewed eight internal investigations and found that 80 percent of the cases were initiated by the department, not by citizen complaints — an indicator the reviewers said tends to show internal attention to potential misconduct.
Teresa Magula, OIR’s lead reviewer on numbers and case details, said the review covered 40 separate allegations across the cases, with 32 sustained findings. "In part, it's what Steve was speaking to earlier. The large percentage of cases were initiated internally," Magula said. She added that the department has been improving the balance and framing of allegations in internal files.
The audit highlighted two critical incidents. One, opened in 2022, saw a protracted administrative investigation that OIR said took three years to complete because the department deferred its administrative probe until related civil litigation closed. OIR recommended conducting administrative and criminal inquiries concurrently when practicable so that public reporting and internal review timelines do not stretch for years.
The second critical incident was an October 2023 fatal officer‑involved shooting near public housing. OIR said there were on‑scene management and evidence‑preservation concerns — including an officer moving evidence before controls were in place — though supervisors intervened. The department completed its administrative investigation in July 2024; OIR held reporting to the council until civil litigation closed. "The department proceeded with charges before the criminal side was completed," Magula said, noting the department took a proactive approach but that litigation timing affected public reporting.
OIR praised the department’s force review cadre — a subject‑matter panel that reviews use‑of‑force incidents and now considers de‑escalation, supervision and duty to intercede — and encouraged expanded, predictable use of that cadre. Reviewers also said each investigated case included training or counseling follow‑up, regardless of final findings, a practice OIR has encouraged.
On technology, OIR noted a new trend: officers using AI tools to draft police reports. Reviewers said the software can improve efficiency but also risks producing overly uniform or artificially generated narratives. Connolly and Magula described guardrails that should be applied, including supervisor review and verification checks: "One of the very basic functions is that it will throw little Easter eggs into the report that it has generated, and it is expecting an officer to capture those in his or her review of the report," Magula said. OIR recommended local controls and supervisor oversight to ensure accuracy and individual officer accountability.
OIR also reported that the department has reduced investigative timelines in recent years, hired a legal adviser to support critical matters and that internal debriefs after critical incidents are occurring to identify training, equipment or tactical needs. The reviewers said they found no statistical evidence in their sample that misconduct or enforcement was disproportionately applied by race or ZIP code, but they cautioned the complaint sample sizes remain small and that quantifying bias requires larger datasets and sustained analysis.
OIR said it will continue community outreach, and announced a public engagement session the reviewers will hold in council chambers; the public session was scheduled for 6–8 p.m. on the evening of the work session. The council did not take action on the audit; the presentation is part of OIR’s regular reporting obligations.

