Honolulu Police Consider AI for Report Writing: Efficiency vs. Integrity Debate

Honolulu Police Department Explores AI-Assisted Report Writing

The Honolulu Police Department (HPD) is considering the implementation of artificial intelligence to assist officers in writing reports, a move that could significantly impact law enforcement practices in Hawaii 1

1

. Interim Police Chief Rade Vanic announced that the department might pilot this program by the end of the year or early spring, utilizing technology provided by Axon, HPD's body camera vendor 1

1

2

2

.

The Promise of Efficiency

Proponents of the AI-assisted report writing system argue that it could substantially increase police efficiency. According to Vanic, officers currently spend a significant portion of their time typing reports, with estimates suggesting that 30-35% of an officer's time is dedicated to report writing 1

1

2

2

. The AI technology, which uses audio transcripts and visuals from body camera footage, aims to generate draft report narratives in seconds, potentially cutting report writing time in half 2

2

.

Vanic emphasized that this technology could allow officers to spend more time on active policing duties:

"Instead of sitting parked on the side of the road spending 30-40 minutes writing up a report, an officer could be driving around his or her beat, talking to people in the community or conducting traffic stops," 1

1

Concerns and Criticisms

Despite the potential benefits, the proposal has faced significant criticism from various stakeholders. The American Civil Liberties Union (ACLU) of Hawaii has expressed strong opposition to the idea. Wookie Kim, legal director at the ACLU of Hawaii, stated:

"AI technology is just too unreliable, untested, biased and also opaque. It's not yet ready to be infused into one of the most important parts of the criminal legal system, which is the police report." 1

1

2

2

Critics highlight several potential issues with AI-generated reports:

1. Reliability concerns: The technology has been known to produce errors and even "hallucinate" information 1

   1

   2

   2

   .
2. Bias: AI systems can inherit biases based on their training data, potentially favoring law enforcement perspectives over public interests 1

   1

   2

   2

   .
3. Lack of context: AI may not understand the nuanced context of situations that human officers can perceive 1

   1

   2

   2

   .
4. Integrity of the criminal justice system: As police reports often serve as the gateway to the legal process, critics argue they are too important to delegate to AI 1

   1

   2

   2

   .

Regulatory Landscape and Future Steps

Currently, only Utah and California require police departments to disclose when AI is used in report writing 1

1

2

2

. California Governor Gavin Newsom recently signed a law mandating such disclosure 1

1

2

2

.

Vanic acknowledged the need for careful consideration and community input before implementing the technology. He stated that HPD would need to develop a comprehensive AI policy and work with stakeholders, including the judiciary and prosecutor's office 1

1

2

2

.

As the debate continues, the Honolulu Police Department faces the challenge of balancing technological innovation with the integrity of its policing practices and the criminal justice system at large.