Current AI/ML solutions in network management rely heavily on ‘black box’ methods with limited explainability, creating a gap between AI capabilities and stakeholder trust. To address this, the “Trustworthy AI for Automating Ultra-Large Network Optimisation” project focuses on integrating Artificial Intelligence (AI) to manage the complex 6G ecosystem.

Objectives

The project’s key research challenge is to define, communicate, and scale multi-stakeholder levels of trust and the necessary AI explanations to develop trustworthy autonomous communication systems that maximize quality of service and user experience.

As AI and machine learning (ML) become more embedded in managing large-scale networks, there’s a crucial need for solutions that improve interpretability, transparency, and explainability. The goal is to translate these properties into quantifiable measures of trustworthiness for autonomous networks.