Project

FAIR-PReSONS

Fair predictions of gender-sensitive recidivism

Timeframe

06/2024 – 05/2026 (24 months)

Get in touch

Developing a bias-free AI system for the fair prediction of recidivism

FAIR-PReSONS project will develop a bias-free AI system for the fair prediction of recidivism, emphasising gender equality perspective, conforming to EU legislation for non-discriminative AI. To do so, FAIR-PReSONS will first conduct a gender analysis, which maps the potential different impact of the project and its activities on women and men as well as girls and boys in all their diversity.

To accomplish these two goals, FAIR-PReSONS will systematically collect related data (e.g., release papers) from prisons and criminal justice organisations, digitise and document it with appropriate metadata/semantics following EU standards.

Where

  • Bulgaria
  • Cyprus
  • Greece
  • Portugal

Funding

  • Justice Programme | Line of funding: JUST-2023-JACC-EJUSTICE

Objectives

Contribute to the understanding of bias-free AI (machine learning) algorithms used for predictions.

Minimize the risk of algorithmic discrimination, in particular in relation to the development of AI systems for supporting the decision-making of judges in recidivism prediction.

Address biases inherent in the data and algorithms used for delinquency and recidivism analysis.

Contribute to the ethical use of AI systems in the justice sector.

Expected outcomes

A bias-free AI system for the fair prediction of recidivism, emphasising gender equality perspective.

Digitalised data from prisons and offence management systems (OMS) from each involved country.

Report on the relevance of gender and other sensitive characteristics in criminal proceedings.

Code of conduct for legal professionals in the use of AI.

Training judges and other legal professionals on the use of the bias-free AI system.

Coordinator & Partners