SAINTS: UKRI AI CDT in Safe Artificial Intelligence Systems

All SAINTS research sits under two themes. Lifelong safety of AI systems is about building safety assurance in from the first design decisions through to use after deployment, because AI systems now work in open and changing settings. Safety of increasingly autonomous AI systems is about what follows when decisions pass from people to machines.

The research is grounded in real systems. The centre gives example use cases: self-driving vehicles, including handovers of control and the legal liability of drivers and manufacturers; autonomous ships and underwater robots; surveillance drones; defence and national security; factory robots working beside people; and AI in healthcare and diagnosis, where trust, explanation and moral and legal responsibility matter as much as accuracy.

As a centre for doctoral training, SAINTS gives the whole cohort a structured programme in four streams: taught courses, practice-based learning, event-based learning, and secondments and exchange. The first six months are weighted towards taught courses and team working, with group and individual assessment. Three courses, in AI, Safe AI, and Law, Ethics and Society, each run as a three-day block of lectures, seminars and group work, because cohort members arrive with very different backgrounds.

A regular seminar series covers research and professional skills, and quarterly workshops let researchers and supervisors present and get feedback from across disciplines. Typically in the second or third year, every researcher takes a mandatory secondment of at least three months, full-time equivalent, with a partner or other external body, built around a use case and designed to feed the PhD; researcher and host then report on its impact. The PhD ends in a thesis of 80,000 words, and the programme does not include a separate master's qualification.

Researchers with home fee status receive a tax-free stipend of 31,000 pounds in 2026-27 under the TechExpert pilot, pro rata for part-time study and usually rising with inflation. International researchers are welcome but receive the standard UKRI minimum stipend, and the number of international places is limited. For everyone, annual tuition fees are paid and there is funding for events and conferences. SAINTS does not normally take self-funded students, so that every researcher has the same experience, and it does not accept people in full-time jobs who want to study part-time.

Applicants need a first-class honours degree, or a 2:1 and a master's, or equivalent, and relevant work experience is taken into account. Degrees in computer science, engineering, mathematics, economics, health sciences, law, philosophy and sociology all fit. Basic programming is expected, and a bootcamp before the start of the year helps successful applicants who need it; practical experience of building AI systems is desirable but not essential. English requirements are IELTS 7.0 with at least 6.5 in each part, or an equivalent test.