# Arkansas Law-Enforcement AI Governance
**Arkansas law-enforcement AI governance** is the collection of human-review, approval, security, audit, records, training, and prohibited-use controls ACIN cites for AI-assisted investigative work.
## How it appears in the corpus
- The produced Act 848 copy requires a policy and an authorized human final decision-maker ([[ACIN AI Policy and CLEST Training Packet]], `GOV-201_Arkansas_Act_848_of_2025_AI_Policy_Authority.pdf`, pp. 4-5).
- The statewide policy requires approved tools, documented human review, auditability, security, and training and restricts surveillance/profiling without authority (`GOV-202_Arkansas_AI_Acceptable_Use_Policy_Final_2026-02-26.pdf`, pp. 1-4).
- ACIN's model policy is a template requiring local legal, privacy, procurement, bias, accuracy, retention, and vendor-risk review (`GOV-002_Arkansas_Model_AI_Policy_TEMPLATE.pdf`, pp. 1-2).
- Peregrine's addendum makes AI/ML outputs decision support subject to human review ([[Peregrine and Carahsoft Executed Contract]], `CON-002_Peregrine_Carahsoft_Executed_Contract_2026-03-17.pdf`, p. 22).
## Notes
The records establish standards and proposed training. They do not include completed system approvals, validation studies, audits, or an externally approved CLEST course.