Research Assistant, AI Safety

AISafety; Oxford; Hybrid; GBP 36,000 - 39,000; Full time

Job summary

University of Oxford in Oxford is recruiting a Research Assistant to work with the OATML group on AI safety, security, and robustness, including foundational research on LLMs and multimodal models. You will support projects on safety alignment, adversarial training and uncertainty quantification.

The appointee will typically hold a Master’s degree (or near completion) with strong programming skills and a track record of publishable research.

Qualifications

• Master's degree (or near completion) required.
• Strong programming skills and ability to produce publishable research.
• Knowledge of foundation models, LLM jailbreaking, safety alignment, adversarial ML or deep learning.

Responsibilities

• Provide research support for AI safety, security, robustness projects.
• Contribute to foundational research on frontier models (LLMs and multimodal models).
• Present proposals and findings at meetings and conferences.

Skills

Foundation models

LLM jailbreaking

Safety alignment

Adversarial machine learning

Deep learning

Programming

Education

Master's degree (or nearing completion)

Job description

Location:Wolfson Building, Parks Road, Oxford, OX1 3QD

One Research Assistant (RA) Position - Generative AI, AI Safety, AI Security

One post tenable immediately, to conclude October 1st 2027 with the possibility of extension depending upon successful project review and project funding. Post offered at Grade 6 (£35,681- £39,424 per annum. This is inclusive of a pensionable Oxford University Weighting of £1,730 per year.).

The Department of Computer Science seeks to employ one Research Assistant (RA) to work within a research group, reporting to Prof Yarin Gal and Dr Lin Li. The postholder will be a member of the Oxford Applied and Theoretical Machine Learning (OATML) research group with responsibility for the provision of research support for projects relating to AI technical safety, security, reliability and robustness, as well as foundational research on frontier models like Large Language Models (LLMs) and multimodal foundation models. This includes topics such as safety alignment, adversarial training, jailbreaking, uncertainty quantification, and AI hallucination.

Candidates will be expected to have a Master’s degree (or be close to completion). They should have knowledge of foundation models (generative AI), LLM jailbreaking, safety alignment, adversarial machine learning, or deep learning, along with strong programming skills. Selection will also consider the ability to produce publishable research, communicate proposals and findings effectively, and represent the research group at meetings and conferences.

All applicants must complete an application form and submit a CV and supporting statement.

The closing date for applications is noon Friday 11th September 2026.

We are a Stonewall Top 100 Employer, Living Wage and Mindful Employer, holding an Athena Swan Bronze Award, HR excellence in Research and Race Equality Charter Bronze Award.

Our staff and students come from all over the world and we proudly promote a friendly and inclusive culture. Diversity is positively encouraged, through diversity groups and champions, for example http://www.cs.ox.ac.uk/aboutus/women-cs-oxford/index.html , as well as a number of family-friendly policies, such as the right to apply for flexible working and support for staff returning from periods of extended absence, for example shared parental leave.

Demonstrating a commitment to provide equality of opportunity. We would particularly welcome applications from women and black and minority ethnic applicants who are currently under-represented within the Computer Science Department. All applicants will be judged on merit, according to the selection criteria.

£35,681 to £39,424 per annum. This is inclusive of a pensionable Oxford University Weighting of £1,730 per year.