Loss of right to liberty and security

Mitigation strategy

1. Implement an absolute prohibition or a highly restrictive set of design constraints on the use of AI systems in contexts where they serve as the sole or primary basis for arrest, detention, or liberty-restricting decisions (e.g., mandatory non-use in initial arrest determination without human confirmation of independent evidence), aligning with the principle of harm avoidance. 2. Institute mandatory, independent human review and judicial authorization for any liberty-depriving action suggested or enabled by an AI system. This review process must explicitly check the AI's output against international human rights standards of legality, necessity, and proportionality, ensuring due process is always observed. 3. Develop and maintain transparent, auditable records for all AI-enabled law enforcement actions, ensuring individuals subject to AI-facilitated arrest or detention have an immediate right to effective remedy, including access to the underlying algorithmic rationale and evidence for the detention, as required by due process and the right to challenge arbitrariness.