Truace
Subscribe
70/100

Moderate evidence

Cite
More
Export CSVExport JSONEmailLinkedInX

Police told AI backs their enforcement turned harsher, less educational, in a 356-officer experiment

Study title: Artificial intelligence, accountability mechanisms, and the transformation of street-level enforcement style: experimental evidence from frontline policing

Evidence type
Randomized controlled trial
Sample
n = 356
Sources
1 sourcePeer-reviewed
Published
2 min read
Truvace record
TRV-2026-2523cite or check it

Study title: Artificial intelligence, accountability mechanisms, and the transformation of street-level enforcement style: experimental evidence from frontline policing

A preregistered survey experiment with a representative sample of 356 street-level police officers tested how AI support changes the way officers handle honking violations. Officers working with AI support leaned toward a more formal and coercive enforcement style, and the educational side of their approach shrank.

Read the study abstract

Artificial intelligence (AI) is increasingly used to support street-level enforcement, yet its impact on bureaucrats’ enforcement style under different accountability mechanisms remains understudied. To achieve this, we conducted a preregistered survey experiment in the context of honking violations with a representative sample of 356 street-level police officers. Results show that AI-supported enforcement shifts bureaucrats toward a more formal and coercive style while diminishing its educational component. Pro…

The problem

In a preregistered survey experiment with 356 street-level police officers handling honking violations, AI-supported enforcement shifted officers toward a more formal and coercive style and reduced the educational part of their approach.

In brief

The study also compared two kinds of accountability. Process accountability led to a more educational, prioritization-focused and accommodative style, while outcome accountability reinforced a formal and coercive style.

The two factors interacted. Under process accountability, AI's effects grew stronger, adding to formalism and coercion and further weakening the educational approach. The findings come from a survey experiment on one type of violation, so they describe stated enforcement choices rather than observed street behaviour.

Main points

  1. The experiment was preregistered and used a representative sample of 356 street-level police officers.
  2. The scenario was honking violations, a low-level traffic offence.
  3. AI-supported enforcement moved officers toward a more formal and coercive style.
  4. AI support reduced the educational component of enforcement.
  5. Process accountability encouraged a more educational, prioritization-focused and accommodative style.
  6. Outcome accountability reinforced a formal and coercive style.
  7. Process accountability amplified AI's effects, adding formalism and coercion and weakening the educational approach.

The full story

The research looks at street-level bureaucrats, here frontline police officers, and how their enforcement style changes when AI supports enforcement. It uses a preregistered survey experiment built around honking violations, with 356 officers in a representative sample.

The main result is that AI support pushes officers toward formal, coercive enforcement and away from educating drivers. This matters because enforcement style shapes how rules are applied in everyday encounters with the public.

Accountability mechanisms changed the picture. Process accountability favoured educational, prioritization-focused and accommodative enforcement on its own, while outcome accountability favoured formal and coercive enforcement. When combined with AI, process accountability strengthened the AI effect instead of offsetting it.

Sources

1 on file
  1. Global Public Policy and Governancedoi:10.1007/s43508-026-00146-y
    Peer-reviewed2026-06-01

Talk it over

Four questions to start onWho carries this cost, and what would have to change to prevent it?

Start with a question

Police told AI backs their enforcement turned harsher, less educational, in a 356-officer experimenthttps://truvace.com/item/artificial-intelligence-accountability-mechanisms-and-the-transformation-of-stre

  1. Who carries this cost, and what would have to change to prevent it?

    The problem In a preregistered survey experiment with 356 street-level police officers handling honking violations, AI-supported enforcement shifted officers toward a more formal and coercive style and reduced the educational part of their approach.

    Answer in the thread

  2. How do we know? What can this kind of evidence show, and what can’t it?

    The evidence This rests on a randomized controlled trial with a sample of 356, from Global Public Policy and Governance.

    Answer in the thread

  3. What does it leave open? What else would you want to know before acting on it?

    The record’s own caveat The evidence comes from a survey experiment on honking violations, so it measures officers' responses to a scenario in one enforcement context rather than real-world enforcement outcomes.

    Answer in the thread

  4. Have you seen AI used this way where you study, work or live? Did it go the same way?

    Answer in the thread

The questions are the same on every record. The details under them are taken from this record as written.

What readers say

More from Crime