Beyond Regulatory Sandboxes: TUM Think Tank at the Privacy Laws & Business International Conference
What does it take to make regulatory sandboxes work? At the Privacy Laws & Business International Conference in Cambridge, Andras Molnar shared the TUM Think Tank’s research on sandbox implementation in a panel with the UK ICO and Singapore’s PDPC.
Andras Molnar, Senior Digital Policy Manager and Director of Online Safety at the TUM Think Tank, spoke at the 39th Privacy Laws & Business International Conference, “Digital Regulation: The fundamental things apply.” The conference brought together senior representatives from the ICO, EDPB, DSIT, the Irish DPC, Singapore’s PDPC, OpenAI and TikTok, as well as leading international law firms and universities, including the University of Cambridge and UCL.
On 7 July, he joined the panel “Beyond regulatory sandboxes to more agile, learning-oriented models.” He spoke alongside Adeline Tung, Director (Policy & Technology) at Singapore’s Personal Data Protection Commission. The session was chaired by William Malcolm, Executive Director for Regulatory Risk and Innovation at the UK ICO. The panel looked at what sandboxes are, how they work in practice, and how they can move beyond single-regulator, single-company setups toward cross-sector models built around learning.
Andras showed that sandboxes are hard to put into practice. They need dedicated resources, long-term commitment from institutions, and coordination across legal, technical, policy and societal areas. He presented three building blocks as a guide to where projects usually run into problems: a clear technical proposal, adequate resources and expertise, and a work plan with checkpoints for learning and managing risk. Adeline Tung shared lessons from Singapore’s PET Sandbox, where a clearly defined policy objective was key to success.
The talk built on the article by Andras Molnar, Urs Gasser, Armando Guío Español, Sandra Cortesi and Markus Siewert in Privacy Laws & Business, “Sandboxes: Tools for regulatory experimentation and learning,” and on the TUM Think Tank policy brief “AI Sandboxes: A Key Tool for Adaptive AI Governance.”