Foundations of Secure Digital Solutions and Artificial Intelligence (TEM-TA119)

General Data

Abstract. The complexity of digital systems and the widespread use of artificial intelligence in all areas of life increase the need to ensure their correct functioning and security. Serious threats arise from software vulnerabilities, programming errors, unreliability of artificial intelligence, and malicious attacks, which can lead to privacy violations and the spread of misinformation. The project has brought together research groups in data science, cryptography, and programming languages to develop new methodologies for enhancing the reliability of digital solutions. The goal is to create a foundation for evidence-based approaches that ensure the correctness and security of digital systems. These methodologies include abstract semantics, automatic verification of software, verification of delegated computations, zero-knowledge proofs, and enhancing the reliability of artificial intelligence.

PIs: Meelis Kull, Helger Lipmaa, Raul Vicente, Vesal Vojdani

Duration: 01.09.2024−31.08.2028

Link in ETIS: https://www.etis.ee/Portal/Projects/Display/32e18985-319d-487b-98a1-1209aad50688

Crypto Group

The promise of the group led by Helger Lipmaa was the following:

Official Description. The cryptography group focuses on developing and implementing zero-knowledge proofs and verifiable computation methods. Existing methods allow the presentation of privacy-preserving ("zero-knowledge") proofs that can be efficiently verified independently of the complexity of the original computation. Unfortunately, proof generation is still inefficient. The working group will investigate the creation of more efficient and secure proofs and their use in different digital solutions and artificial intelligence.

Unofficially, we plan to study efficiency and security of zk-SNARKs, with zkML being one of the focus areas.