PhD Studentship: Cyber and Privacy Risk Management of AI systems

AI has taken its rightful place as a key disruptive technology over the last few years. The availability of open-source software platforms (e.g. Hugging Face, OpenAI, VertexAI) enables the sharing and access of AI models and their source code. Although these platforms make accessibility to AI models more ubiquitous, they often include significant vulnerabilities and given that the perverseness of AI systems is inevitable, there is a need for sustainable methodologies to comprehensively identify new vulnerabilities and ways of modelling threats throughout the AI lifecycle. This PhD aims to develop a sustainable framework to support the deployment of secure, privacyenhanced, and transparent AI systems by addressing the following research objectives. Objectives:

• develop threat modelling approaches to identify new vulnerabilities and attacks on AI systems.
• systematically identify and quantify cyber and privacy risks associated with AI systems. This involves identifying controls to mitigate the identified risks.
• develop sustainable approaches to fortify AI systems against evolving threats considering the cost and benefits of different controls.

The outcomes will cultivate a safer, reliable, transparent, and inclusive environment for the use of AI while reflecting its profound importance to society. The PhD candidate will work under the primary supervision of Dr Sakshyam Panda at the Centre for Sustainable Cyber Security (CS2) (https://www.gre.ac.uk/research/groups/sustainable-cyber-security-cs2 ), University of Greenwich, London. The supervisory team includes Professor Emmanouil (Manos) Panaousis and Professor George Loukas.

The University of Greenwich (through CS2) has been recently recognised by the UK government as a NCSC Academic Centre of Excellence in Cyber Security Research (https://www.ncsc.gov.uk/information/academiccentres-excellence-cyber-security-research). CS2 has a vibrant research community with 35 permanent members of staff and 20 PhD students in a wide range of cyber security areas. As part of our research team, the candidate will have access to state-of-the-art research facilities and work with a highly collaborative and interdisciplinary team of researchers Diane Gan PhD Studentship – Centre for Sustainable Cyber Security – 2024 from different fields including cyber security (including human-centric cyber security), security and privacy of AI, machine learning, formal method, natural language processing and more.

Advert information