Europol calls for regular training, testing of models and consistent frameworks to avoid AI bias in law enforcement
INNOVATION: A new report from Europol which offers practical guidance on tackling artificial intelligence (AI) bias in law enforcement is calling for more regular training for personnel “to deepen their understanding of AI technologies”, routine and regular multi-stakeholder reviews of AI models, and more consistent frameworks around their use, to combat the dangers of AI bias – particularly in areas such as predictive policing – as Policing Insight’s Sarah Gibbons reports.