The seven research objectives (SOs)
SO1 – Accurate AI
Develop AI systems that produce reliable, precise results across different contexts while addressing ethical and legal trade-offs between accuracy, interpretability, privacy, and sustainability.
SO2 – Interpretable, Transparent AI
Advance explainable AI (XAI) by improving the comprehensibility and scalability of explanations, ensuring that AI systems’ decision-making processes are transparent and accountable.
SO3 – Aligned, Fair, Inclusive, and Autonomy-Protecting AI
Ensure AI goals and behaviours align with human values, fairness, inclusiveness, and respect for autonomy, developing adaptive approaches to handle contested ethical standards and dynamic systems.
SO4 – Safe AI
Design AI that operates safely and without harm, taking a systems-level and real-time perspective on how humans and institutions use and trust AI in practice.
SO5 – Sustainable AI
Make AI less resource-intensive by improving energy efficiency, reducing environmental impact, and contributing to the green transition through “green AI” and governance innovations.
SO6 – Democratic Governance of AI
Develop participatory governance mechanisms that ensure accountability, inclusion, and citizen engagement in AI development and deployment, addressing current deficiencies in global AI regulation.
SO7 – AI at Scale
Tackle challenges in deploying AI with vast data volumes, distributed systems, and real-time use—ensuring that large-scale AI remains accurate, safe, transparent, and aligned.
Together, these objectives form the foundation for TRUST’s 14 Research Areas and 15 Action Clusters, integrating perspectives from computational and data science, law and social science, and philosophy to make AI systems ethically, legally, and technically trustworthy.