AI Machine Learning and Regulation: The Case of Automated Vehicles

From road to rail to shipping, recent technology is leading to advances in automated vehicles: driverless cars, trains and boats. Known as “AVs”, they promise improved safety and accessibility. But they are also cause for concern. Potential risks are linked to data quality and representation, the development and verification of AI models, increased vehicle travel, land-use impacts, and deskilling of vehicle operators.

To safely harness the full potential of automated transport, supporting regulation must be able to demonstrate AVs trustworthiness, both in terms of safety and ability to serve the collective good.

This report examines the regulatory approaches to address these challenges – in particular focusing on road vehicles. It provides a common understanding of AI-based automated transport systems and the principles that should form the basis of institutional and regulatory actions to increase safety and social acceptability.