10 Security and Privacy Problems in Self-Supervised Learning
share
Self-supervised learning has achieved revolutionary progress in the past several years and is commonly believed to be a promising approach for general-purpose AI. In particular, self-supervised learning aims to pre-train an encoder using a large amount of unlabeled data. The pre-trained encoder is like an "operating system" of the AI ecosystem. Specifically, the encoder can be used as a feature extractor for many downstream tasks with little or no labeled training data. Existing studies on self-supervised learning mainly focused on pre-training a better encoder to improve its performance on downstream tasks in non-adversarial settings, leaving its security and privacy in adversarial settings largely unexplored. A security or privacy issue of a pre-trained encoder leads to a single point of failure for the AI ecosystem. In this book chapter, we discuss 10 basic security and privacy problems for the pre-trained encoders in self-supervised learning, including six confidentiality problems, three integrity problems, and one availability problem. For each problem, we discuss potential opportunities and challenges. We hope our book chapter will inspire future research on the security and privacy of self-supervised learning.
READ FULL TEXT