About this role
Summary: Lead theoretical research on combining evaluation evidence for machine-learning systems within the Trustworthy and Assured AI theme. Develop formal accounts of evidence combination, contribute to evaluation studies, and provide intellectual leadership for the research project while collaborating with stakeholders...
The Alan Turing Institute's Fundamental Research mission exists to make the advances in data science and AI on which the Institute's other missions and the wider field depend: new methods, new theory, and the foundations that let AI systems be built and used with confidence. Within it, the Trustworthy and Assured AI theme asks how we can know that an AI system deserves the trust placed in it, exploring questions such as how to evaluate what a system does, how to turn evaluation results into assurance-grade evidence, and how to assemble that evidence into an argument that a designer, regulator or operator can act on. The theme works across domains, with a current emphasis on national security and resilient critical national infrastructure. It also maintains the Turing’s Trustworthy and Ethical Assurance (TEA) platform - an open-source tool for building and sharing structured assurance arguments.
This role sits within the Turing's Fundamental Research mission and the Trustworthy and Assured AI theme. The post holder will lead the theoretical strand of the partnership's work, contribute to its evaluation studies and groundwork, and help shape the theme's longer-term research on AI evaluation and verification.
ROLE PURPOSE
The Research Associate will lead a scoped research project on the combination of evaluation evidence about machine-learning systems. The project addresses a gap in current assurance practice: evaluation and explanation methods each produce a result, but a result is not yet a claim. Whether two results count as two pieces of evidence or one repeated depends on whether the methods can fail independently, and no settled account exists of how such results combine, how their reliability should discount them, or what the combined verdict licenses a designer to claim. The post holder will develop that account, drawing on evidence theory, probabilistic reasoning and current work on machine reasoning and human interpretation, and illustrate it on results from the theme's evaluation platform. Alongside the theoretical work, they will contribute to the partnership's evaluation studies, to the mapping of evaluation evidence onto defence and civil-aviation assurance frameworks, and to the design of protocols for eliciting operational requirements from domain experts.
Internally you will be expected to collaborate with colleagues across Fundamental Research and other missions (e.g. Defence and National Security). Externally, the main stakeholders are, initially, the industry partner and, through it, government stakeholders in defence and national security. The role involves working in a defence and national security context and will require security vetting.
HOW YOU WILL MAKE AN IMPACT
- Publish and disseminate high-quality research papers and publications detailing research outputs and project case-studies.
- Cultivate strong relationships and work collaboratively with internal stakeholders, academic experts and research partners across the Turing and wider Turing community.
- Communicate and present research outputs to diverse stakeholders, including project partners, through conferences, events, sprint reviews, workshops, meetings and press opportunities as appropriate.
- Adhere to and promote principles of reproducible and ethical data science and ensure secure handling of data and health and safety in all aspects of work.
- Undertake high-quality research, contributing to the evaluation and assurance of machine-learning systems within the Trustworthy and Assured AI theme and Fundamental Research.
- Lead the research and writing of a peer-reviewed paper setting out a formal account of how evaluation and explanation evidence about machine-learning systems can be combined, illustrated with results from the theme's evaluation platform.
- Draft specifications and handling plans for datasets curated with partners, and contribute to the design of the theme's evaluation platform.
- Contribute to the theme's longer-term research on AI evaluation and verification, including evaluation methodology for agentic and language-model systems.
- Work within the security requirements of a defence and national security context, including the handling of partner material.
ADDITIONAL DUTIES FOR SENIOR RESEARCH ASSOCIATE LEVEL
- Provide intellectual and technical leadership for the research project, defining and refining the research direction with the Principal Investigator and wider team.
- Lead engagement with project partners and stakeholders on research priorities, assurance requirements and the interpretation of evaluation evidence, deputising for the Principal Investigator where appropriate.
- Lead the preparation of proposals, applications and reports for follow-on funding and partnership activity.
- Provide leadership, support and mentorship to junior researchers, interns, students or other early-career colleagues where required.
ABOUT YOU
- A PhD (or equivalent experience and/or qualifications) in a relevant area (e.g. computer science, machine learning, statistics, formal methods, or philosophy of science with demonstrable technical skills in programming and machine learning). Equivalent experience may include substantial industry experience in the evaluation, testing or assurance of AI/ML systems.
- Research experience in one or more of: evaluation of machine-learning systems, explainable AI, formal reasoning under uncertainty (probabilistic or Bayesian methods, evidence theory), or AI assurance.
- Experience designing and running evaluations of machine-learning models, including choice of metrics, experimental design and statistical analysis.
- Experience with quantitative data analysis and competency in Python and the scientific machine-learning stack.
- Experience with Git and GitHub for version control and collaborative development, and experience of working on reproducible research code with software engineers.
- Evidence of high-quality publication(s) in a relevant field (published or in-press) commensurate with your career stage
ADDITIONAL REQUIREMENTS FOR SENIOR RESEARCH ASSOCIATE LEVEL
- Demonstrable experience of leading research projects
- Ability to carry out research independently and take the lead on writing up findings as they emerge
- Ability to take the lead on research direction in collaboration with the PIs
Pleasesee the Job Description for a full breakdown of the role and person specification.
APPLICATION PROCEDURE
If you are interested in this opportunity, please click the apply button below. You will need to register on the applicant portal and complete the application form including your CV and covering letter. If you have questions about the role or would like to apply using a different format, please contact us at recruitment@turing.ac.uk.
TERMS AND CONDITIONS
This full-time post is offered, on a permanent basis. The annual salary range is £46,415 - £52,494 for Research Associate level / £57,000 - £59,000 for Senior Research Associate level. Our offers include excellent benefits, including flexible working and family friendly policies, [which can be seen here.](https://www.turing.ac.uk/sites/default/files/2024-10/turing_benefits_guide_2024.pdf)
Candidates who have not yet been officially awarded their PhD will be appointed as Research Assistant at a salary of £43,445 (pro-rated to working hours) per annum.
Eligibility for Security Check (SC) clearance is a requirement for this role. Eligibility criteria and further information on the process can be found on the UK Government security vetting website.
The Alan Turing Institute is based at the British Library, in the heart of London’s Knowledge Quarter. We offer a flexible, hybrid approach to working, recognising the benefits of both remote working and in-person collaboration. Office attendance expectations may vary depending on the role, team and organisational needs, and may evolve over time. The hiring team will be happy to discuss the current arrangements during the interview process.
EQUALITY, DIVERSITY AND INCLUSION
We value diversity of background, experience, and perspective, and are proud to be an inclusive employer. We warmly encourage applications from all backgrounds, particularly from groups currently under-represented in our sector. If you feel passionate about this role but don’t meet every single requirement, please apply – we recognise that great candidates may bring strengths beyond the criteria listed.
We are committed to making sure our recruitment process is accessible and inclusive. This includes making reasonable adjustments for candidates who have a disability or long-term condition. Please contact us at recruitment@turing.ac.uk to advise us how we can assist you.
Please note all offers of employment are subject to obtaining and retaining the right to work in the UK and satisfactory pre-employment security screening which includes a DBS Check. Full details on the pre-employment screening process can be requested from[HR@turing.ac.uk](mailto:HR@turing.ac.uk.)
Benefits
The Alan Turing Institute is the UK’s national institute for AI and data science.
We are named in honour of Alan Turing, whose pioneering work in mathematics, engineering and computing helped shape the development of modern AI and remains an enduring example of how technical innovation can serve the national interest.
Our mission is to secure the UK’s resilience for the age of AI. We do this by bringing together expertise from across the country to protect critical systems against disruption, counter AI-enabled threats and shape UK strategic AI capability.
Please find more information about us [here](http://www.turing.ac.uk/about-us "http://www.turing.ac.uk/about-us").