Assessing Societal Risks of Online Platforms

The ARTAI (Assessing Risk for Trustworthy AI) project addresses the societal challenge of mitigating risks posed by recommender algorithms embedded in online platforms such as Instagram, TikTok and YouTube. Research suggests that recommender algorithms can have a profound influence on societal views and behaviour, increasing political polarization, negatively impacting youth mental health and reinforcing harmful gender and socio-cultural stereotypes. New AI regulation is being implemented across the EU to mitigate these risks through auditing and research. However, there are currently no tools to enable implementation of these regulations and ARTAI seeks to address this challenge.