Artificial intelligence courses in UK universities could be particularly exposed to any turbulence in applications from European Union students caused by Brexit, an analysis by?Times Higher Education?suggests.

Data from the Higher Education Statistics Agency on course numbers by individual subject show that students from other EU countries made up almost a quarter of the 1,100 on AI undergraduate and postgraduate programmes in 2017-18.

It represented the highest proportion of EU students for any subject where more than 1,000 students were enrolled. A further 30 per cent of the students were from non-EU countries, meaning that more than half of the students were from outside the UK.

EU students starting courses at English universities this autumn will be charged the same fees as domestic students, and will still have access to state-backed loans,?but they may lose this status from 2021-22. A 2017 study??the average impact of EU students facing the same fees as non-EU students and losing access to loans could be a 57 per cent drop in EU enrolments.

ÁñÁ«ÊÓƵ

For artificial intelligence courses, such a drop would potentially represent the loss of dozens of students in a subject area that the UK government?has earmarked?as a key focus after the country leaves the EU.

Bill Mitchell, director of policy at the British Computer Society, the learned society and professional body for IT, said because the UK was recognised as a ¡°world leader¡± in AI development, courses were in huge demand from overseas.

ÁñÁ«ÊÓƵ

He pointed in particular to master¡¯s courses, where again the Hesa data show more than half of students are from outside the UK and a large proportion (17 per cent) from other EU countries.

¡°What will matter after Brexit is how welcoming we remain to the best and brightest overseas applicants and how well we nurture our own home-grown talent,¡± he said.

¡°AI is a global enterprise and UK students will be better technically and ethically as future professionals if they learn as part of a multicultural, diverse cohort of students from around the world.¡±?

Dr Mitchell added that one way to boost the attractiveness of AI university programmes would be to increase awareness of master¡¯s courses that have been developed,?, to enable students taking non-IT subjects to convert to AI or data science.

ÁñÁ«ÊÓƵ

¡°We would recommend the government increases promotion of MSc conversion courses in these areas to EU students currently studying, for example, social sciences, given the prestige of AI in the UK, and the growing global employability of highly skilled graduates in subjects like data science,¡± he said.

Artificial intelligence is not the only specific IT discipline to have a large proportion of students from other EU countries: in bioinformatics, they represented 24 per cent of students in 2017-18 and in computer generated imagery it was 19 per cent.

For courses with the general label of computer science ¨C which involved some 63,000 students ¨C the proportion from other EU countries was 10 per cent.

simon.baker@timeshighereducation.com