Objective

Develop secure and reliable AI solutions that protect IoE infrastructures against data tampering, adversarial attacks, and privacy leakage.

Lead: Vienna University of Technology (TUW) · Timing: M7-M42 · Main ESR involvement: ESR4-ESR6

Key Tasks

  • Design a scalable Blockchain-based secure data sharing framework for heterogeneous IoE entities using lightweight consensus.
  • Develop adversarial machine learning defences against model poisoning and evasion attacks during training and inference.
  • Create a two-level privacy protection mechanism combining parameter pruning and crafted noise generation for responsible AI.

Programme Links

This work package contributes primarily to RO-2 and is part of the proposal-stage programme structure captured in the SAILING Part B document.

Overview

WP2 at a Glance

Lead, timing, and role in the overall SAILING architecture.