PAIR Seed Fund Project Name:
From Code is Law to Law is Code: Towards Intelligent Regulation of Artificial Intelligence
Objectives:
Algorithmic flaws and egregious misuse of artificial intelligence (AI) in the domain of autonomous driving place the welfare of humanity at an unforeseeable stake. With the escalating risk-profile of AI, ex post enforcement measures, resources, and powers envisaged by regulators are unlikely to fulfil necessary regulatory duties.
This project aims to provide a comprehensive technical roadmap for legally-compliant artificial intelligence systems. The team developed a legally compliant autonomous driving program as a prototype. It involved transforming sophisticated black-letter laws into AI/machine learning (ML)-literate parameters, and subsequently training AI/ML models for algorithmic auditing against algorithmic risks, and applying both the symbolic AI approach and neural network deep learning to build AI/ML applications on key applicable AI laws and guidelines.
Key Research Outcomes:
- Published two papers based on the research results, and one of which recognised with the Best Paper Award at the 2023 IEEE International Conference on Intelligent System and Knowledge Engineering
- Submitted proposals to external funding schemes including: Theme-based Research Scheme (TRS), Research Impact Fund (RIF), Research, Academic, and Industry Sectors One-plus (RAISe+)
- Delivered keynote speeches featuring research outcomes at two international conferences
