DSI hosts first Privacy in Machine Learning Meetup at Imperial

On 4 February 2025, the DSI’s Computational Privacy Group hosted a meetup for privacy experts working in machine learning.

The sold-out event saw over 50 privacy experts from across London and beyond meet to discuss the latest developments in machine learning and to meet other like-minded individuals.

The Computational Privacy Group at Imperial’s Data Science Institute set up the event to recognise the growing community of researchers in and around London working at the intersection of privacy and machine learning.

During the meet up which took place at Imperial, privacy experts attended a series of short research talks from a diverse group of researchers, addressing multiple aspects of privacy in machine learning.

Full details on the speakers and their talks can be found below:

• Graham Cormode, University of Warwick, Meta AI – Federated Computation for Private Data Analysis
• Lukas Wutschitz, M365 Research, Microsoft – Empirical privacy risk estimation in LLMs
• Jamie Hayes, Google DeepMind – Stealing User Prompts from Mixture-of-Experts models
• Ilia Shumailov, Google DeepMind – What does it mean to operationalise privacy?

For more information about the speakers and the event itself please visit the event webpage.