Responsible AI Group at the UK Artificial Intelligence Research Symposium (UKAIRS) 2025

Held at Northumbria University on 8–9 September 2025, UKAIRS brought together experts from computer science, engineering, social sciences, and the humanities to shape UK research priorities, encourage cross-disciplinary collaboration, and guide the national conversation on delivering AI that benefits society and the economy.

The Responsible Artificial Intelligence Group from our school participated with a delegation headed by Dr Koorosh Aslansefat, supported by Dr Rameez Kureshi. Across two days, our researchers presented cutting-edge work supported by the National Edge AI Hub, including:

• Mostafa Anoosha – EdgeAI-Enabled Responsible Hybrid Federated Learning

• Zeinab Dehghani – gSMILE: A Novel Method for LLM Explainability and Generative AI using SMILE.

• Parvin Ghaffarzadeh – Move Smart: Monitoring Everyday Activities and Health with Wearable Technology and EdgeAI

Visitors stopped by our demo to learn more about responsible, trustworthy, and edge AI.

National Edge AI Hub Workshop on AI Safety and Security

Our school is part of the EPSRC National Edge AI Hub (https://edgeaihub.co.uk/) (UoH PI: Professor Dhaval Thakker). 

Supported by the Hub, a one-day workshop on AI Safety and Security was co-organised by Newcastle University and the University of Hull. The workshop was co-chaired by Dr. Koorosh Aslansefat (Hull) and Dr. Varun Ojha (Newcastle), and it was formally opened by Prof. Dhaval Thakker (Hull). Topics and presentations delivered by UoH researchers included:

AI Explainability 

• Zeinab Dehghani: LLM Explainability

• Mohadeseh Mollapour: Captioning Explainability

Safety and Assurance 

• David Moyneux: LLM Assurance

Federated Learning 

• Kuniko Paxton: Fair Pruning in Federated Learning

• Mostafa Anoosha: Hybrid Vertical Federated Learning