Data and AI Regulation: Pathways to regulatory Sandboxes
The rapid development of the AI technologies is widely associated with the ever-growing amount of data and its increased accessibility. Looking into the future, data is also commonly regarded as being of paramount importance to the evolution of AI. Responsible AI as a foundational requirement, criticality of AI lifecycle data, rise of synthetic data and continuous model evaluation and tuning are recognised as top trends.
These innovations require a safe and reliable regulatory and legislative landscape which balances innovation with the societal interests. AI regulations around the world acknowledge the need for agile anticipatory regulation and various tools and mechanisms are being implemented. Regulatory sandboxes play a pivotal role in creating a new better regulation for AI which is recognised on EU level through their inclusion in the draft AI Act. What is unique for the EU approach is the attempted coordination between implementation of the regulatory sandboxes on local (Member States and regions) and transnational (bilateral, multilateral, EU) level. This session aims to focus on the different levels of regulatory sandboxes with specific highlight on the results challenges and opportunities of local initiatives. It is further going to discuss the scope of the experimentation in regulatory sandboxes and the role of data and societal requirements in the process.
Presentations of the session:
Jonathan Middleton presentation