Logo

Secure computing: computing on encrypted or private data

Xavier Leroy

Collège de France, chair of Software sciences, 2025-2026

Abstract

Cryptography (encryption, signatures, etc.) ensures the confidentiality and integrity of data when it is at rest (stored in files and databases) or in transit (over networks). However, in order to perform calculations on this data, it must traditionally be decrypted and revealed in its entirety to the operator performing the calculation, with the risk of data leaks that this entails. Recent advances in cryptography and cryptographic protocols have led to other approaches in which data remains encrypted or private during calculation, thus guaranteeing its confidentiality. This course is an introduction to these approaches to secure computation. It presents recent results in this area: secure multi-party computation, homomorphic encryption, zero-knowledge proofs, cryptographic obfuscation, etc., as well as some of their applications, such as electronic voting.

Lectures

Video recordings (in French) are available on the Collège de France site.

  1. Secure computing: introduction and case studies.
  2. Fully homomorphic encryption: computing on encrypted data (part 1).
  3. Fully homomorphic encryption: computing on encrypted data (part 2).
  4. Secure multi-party computation: secret sharing.
  5. Secure multi-party computation: garbled circuits and oblivious transfert.
  6. Zero-knowledge proofs.
  7. Secure computing: other directions and conclusions.

Seminar talks

Video recordings and PDF slides are available on the Collège de France site.


Xavier.Leroy@college-de-france.fr