Artificial Intelligence In Law Enforcement And Criminal Justice System: Transparency And Civil Participation As Core Issues In Promoting Human Rights

In this policy brief, we analyze what to consider for mitigating risks of Artificial Intelligence (AI) use in law enforcement (LE), also assessing whether this technology is necessary or appropriate in each context. We focus on how improving transparency, public participation, and accountability are crucial to this task, especially involving the communities most affected by Al-based mass surveillance. Al-based systems deployment in LE can affect fundamental rights like freedom of movement, speech, assembly, privacy, and data protection. A key concern is that Al in LE reproduces historical discrimination based, for instance, on race, gender, and class. Facial recognition technology (FRT) is an example of how systemic racism is ingrained in digital tools. Studies have shown that FRT has a high risk of error in non-white groups. It is also necessary to consider the impacts ofAl-based decision-making technology on the presumption of innocence and due process when used in the criminal justice system. Lack of transparency and civil participation surrounding government adoption ofAl systems by LE aggravate these issues. Another concern is the limits imposed on public scrutiny and oversight over these tech initiatives, with obscurity becoming a feature of their implementation. Policymaking on LE use of Al must ensure proper means of civil participation, especially by the groups mainly targeted by state mass surveillance. We relied on research evidence to elaborate on why governments must include these actors in policy decisions involving adopting Al for security purposes. We also discuss ways to enable civil oversight and its role in providing accountability for security agencies using Al. This subject is essential to the G20 agenda given the global nature of Al LE industry and its relevance for international cooperation.