Research Associate in Testing AI Ethics

Job Description

Within the  UKRI Trustworthy Autonomous System Node on Verifiability, we have an exciting opening for a postdoctoral research on testing ethical aspects of AI. These include testing aspects such as fairness and bias, as well as privacy and explainability. You will be working within a vigorous multi-disciplinary environment of the  Trustworthy Autonomous System program with researchers across many disciplines designing and verifying the autonomous systems of the future. You will be using and advancing the state of the art testing (validation and verification) methods to provide assurances of ethical decision making in sub-symbolic AI components, such as neural networks and statistical machine learning techniques.   

This post will be offered on  a fixed-term contract for 18 months.

This is a full-time  post - 100% full time equivalent.

Key responsibilities

• Developing world-leading research results in collaboration with the team, on testing ethical concerns on AI;   
• Developing a comprehensive and rigorous set of requirements for such ethical concerns;  
• Developing public benchmarks and datasets for evaluating the developed techniques and their future alternatives;
• Coordinating the related activities within the team and within the larger project and program.

The above list of responsibilities may not be exhaustive, and the post holder will be required to undertake such tasks and responsibilities as may reasonably be expected within the scope and grading of the post. 

Skills, knowledge, and experience

Essential criteria  

1. A doctoral degree in Computer Science, Software Engineering or closely related fields (also candidates with their dissertations submitted or under submission will be considered).  
2. Solid track record of study and research in software engineering and software testing (validation and verification) 
3. Proficiency in programming; familiarity with AI frameworks  
4. Possessing analytical skills; basic knowledge of statistical methods 

Desirable criteria

1. Affinity with ethical theories and their formalisation  
2. Affinity with formal verification and mathematical modelling techniques