Apple has released four recordings and a summary of its 2026 Workshop centered on Privacy-Preserving Machine Learning and AI. The event lasted two days and included discussions from Apple researchers and other experts on various topics like Private Learning, Statistics, and security challenges in machine learning.
The workshop featured presentations such as ‘Crypto for DP and DP for Crypto’ by Apple Research Scientist Kunal Talwar. Other notable talks included contributions from Aleksandar Nikolov of the University of Toronto on Online Matrix Factorization, Elissa Redmiles from Georgetown discussing responsible data collection, and Franziska Boenisch from CISPA on mitigating memorization in foundation models.
Apple also acknowledged 24 works presented during the workshop, including three significant papers developed by its researchers, addressing topics like machine learning and homomorphic encryption and privacy loss accounting. For more details and to view the sessions, users can follow the provided link in Apple’s blog post.