Circles Graphics

BLOGS

AI and data science degrees are quietly reshaping UK student housing demand
September 22, 2026

AI and data science degrees are quietly reshaping UK student housing demand

A new demand pattern hiding inside old numbers

UK student housing has long been analysed at a national level — total student numbers versus total bed count, with a "shortage" headline attached. That framing is increasingly unhelpful.

UCAS and HESA data over recent admissions cycles point to a structural shift: computer science, AI, data science and health-adjacent courses (nursing, allied health, biomedical science) have grown their share of UK undergraduate and postgraduate intake, while some traditional humanities and business pathways have plateaued or contracted. This isn't evenly spread. It is concentrated in a specific set of universities — often research-intensive institutions with strong STEM and health faculties, located in cities that already have tight rental markets.

For developers, lenders and investors in purpose-built student accommodation (PBSA), the practical question isn't "how many students nationally" — it's "which cities are absorbing this course-mix shift, and do their accommodation pipelines reflect it." REalyse's planning and demographic data suggests, in a growing number of cases, the answer is no.

Where course growth is outpacing bed delivery

Postgraduate and conversion-heavy AI/data courses tend to cluster in a recognisable set of cities: London, Manchester, Edinburgh, Bristol, Birmingham and a handful of cities with strong health science faculties such as Leeds, Sheffield and Newcastle. Many of these are also cities where PBSA pipeline growth has cooled relative to the mid-2010s boom, as land values, construction costs and planning constraints have pushed some operators toward lower-risk regional cities instead.

REalyse's planning application data — tracked by scheme description, unit count, planning stage and local authority — allows this mismatch to be mapped directly. In several of the cities seeing the strongest course-mix shift toward AI, data and health disciplines, active PBSA schemes in the pipeline (granted or under construction) represent a materially smaller proportion of enrolled full-time students than in cities with more balanced course growth. That gap matters most for postgraduate-heavy intakes, since postgraduate students are typically underserved by traditional PBSA products built around undergraduate-style cluster flats.

The practical effect: rental demand is being pushed back into the private rented sector and HMO stock in exactly the cities where that stock is under the most regulatory pressure.

Why postgraduate-heavy course growth changes the accommodation calculus

AI, data and advanced health courses skew disproportionately toward one-year postgraduate conversion programmes and taught masters intakes, cohorts that behave differently from the standard three-year undergraduate population PBSA has historically been built for. They arrive later in the letting cycle, often need shorter or more flexible tenancies, and are more price-sensitive to premium PBSA rents relative to their funding structure.

This tends to route demand toward two channels: studio-led PBSA (where available) and licensed HMOs near campus and transport nodes. Both are constrained in different ways — studio-led product is expensive to deliver at scale, and HMO stock is contracting in several university cities as selective and additional licensing schemes tighten.

HMO licensing is quietly tightening supply where demand is rising

Several local authorities with large student populations — including parts of London, Manchester, Nottingham and Leeds — have introduced or extended Article 4 directions and additional/selective HMO licensing in the past few licensing cycles, restricting new HMO conversions and, in some cases, capping licence renewals in saturated wards.

REalyse's HMO licensing data, tracked by council and postcode alongside licensed capacity, shows this tightening is not uniform. It is often most aggressive precisely in the wards adjacent to universities expanding AI, data and health faculties — a combination that squeezes the traditional "release valve" for student demand overflow at the same time as course-driven demand is rising.

For lenders and investors, this has underwriting implications. Rental growth assumptions for PBSA schemes in these markets may be more defensible than headline city-level averages suggest, because a shrinking HMO alternative reduces competitive downward pressure on PBSA rents. Conversely, HMO-focused portfolios in the same wards face licence renewal risk that should be priced into acquisition yields, not treated as a background regulatory footnote.

Reading the signals before committing capital

The cities worth watching are not necessarily the ones with the largest student populations, but those where three trends intersect: strong recent growth in AI/data/health course enrolment, a PBSA pipeline that hasn't kept pace in planning applications and granted units, and an HMO licensing regime moving toward greater restriction rather than away from it.

Comparable analysis — cross-referencing planning pipeline by scheme type and local authority against rental listing trends and licensed HMO capacity — offers a more granular read than national enrolment statistics alone. It can flag, city by city, whether current PBSA rents and achieved yields reflect a genuine supply gap or simply track wider market sentiment.

Outlook

Course-mix change is a slower-moving variable than interest rates or planning policy, but it compounds. If AI, data and health faculties continue expanding their share of UK admissions, the cities hosting them will keep absorbing disproportionate housing pressure, regardless of whether their PBSA pipelines are built to match. Developers underwriting new schemes, and lenders assessing existing PBSA or HMO-heavy portfolios, have a window to reprice this risk before it becomes consensus. Cross-referencing planning pipeline, licensing trends and rental performance at postcode level — rather than relying on city-wide averages — is likely to be the difference between capturing that mispricing and missing it.

More from Our Research Based on Your Interest