GCSEs and A-levels in the UK are still sat almost entirely on paper as of 2026. Exam boards have announced phased plans to introduce on-screen options in specific subjects, but several of the original start dates have already slipped, and full-scale rollout across most subjects isn't expected before 2030 on the boards' own timelines — and even that depends on regulatory approval from Ofqual that hasn't yet been given. Estonia is further ahead, digitising its state exams subject by subject, but even there the transition is staged rather than complete.
Digital assessment is one of the areas where it's easiest for an article to drift from "what's been announced" into "what's already happened" — so every date below is stated as an announcement or a plan, not a completed fact, and worth re-checking against the relevant board's own site before relying on it for a specific child's exam year.
Where the UK actually stands
Any move to on-screen national exams in England, Wales or Northern Ireland requires sign-off from Ofqual, the regulator — a board can't simply decide to digitise a GCSE and start doing so. Ofqual itself has indicated that on-screen exams could be introduced more broadly by around 2030, framing it explicitly as a multi-year transition rather than a near-term switch.
Individual boards have announced their own more specific, subject-level plans, several of which have already been pushed back:
- AQA proposed the reading and listening components of GCSE Italian and Polish as an early candidate for digital exams, subject to Ofqual's approval, as part of a longer-term ambition for students to be sitting at least one major subject on-screen by around 2030.
- Pearson Edexcel had planned to offer GCSE English Language and English Literature on-screen as an option from summer 2025; that timeline slipped, and Pearson has instead pointed to History and Business as subjects it hopes to bring on-screen from around 2027.
- OCR had been developing a digitally-assessed Computer Science GCSE intended for first teaching in 2025, but delayed the launch, citing factors including the change of government and the incoming curriculum and assessment review.
The pattern across all three boards is the same: a specific, named pilot subject or component, a target date, and — so far, in every publicly reported case — that date moving later rather than earlier. That's worth bearing in mind for any coverage (including future updates to this article) that states a rollout as settled.
Estonia: further along, still not finished
Estonia is regularly cited as a country ahead of the curve on exam digitisation, and the underlying reporting bears that out, with an important caveat: it's further along, not finished. Estonia's exams authority has been moving its state exams onto a digital Exam Information System (EIS), with language exams (Estonian and English) among the first subjects to move to a fully electronic format, and other subjects — including secondary-school-leaving exams — following in stages after that. Reporting from Estonia's national broadcaster ERR notes that even the maths exam for basic-school students remained paper-based as recently as 2025, illustrating that "Estonia has gone digital" is a rolling process, not a single cutover date.
One incidental benefit reported alongside the transition: a digital format can test skills a paper exam structurally cannot, such as listening comprehension delivered as an audio component within the same digital exam session — a capability, rather than the primary motivation, of the switch.
A third data point: the Netherlands
The Netherlands offers a useful middle case between "still mostly paper" (England) and "furthest along" (Estonia). Dutch students already sit part of their final exams under a structure that separates a nationally set centraal examen from a school-run schoolexamen (see UK exam boards vs the Netherlands for the full split), and CvTE, the Dutch central examination board, has explored computer-based delivery for parts of that national component, alongside separately organised digital and practical assessment routes on some vocational (VMBO) programmes. The Dutch approach illustrates a pattern worth naming: countries with an existing split between a centrally set component and a locally run one have a natural place to pilot digital delivery — the centrally set half — because that's the part one body already controls end-to-end, rather than needing to coordinate the change across every individual school's own assessment practice.
Why this is harder than it looks
The cases above illustrate the same underlying difficulty. Converting a high-stakes national exam to computer isn't just a format change — it raises questions about equal access to hardware across schools, accessibility (does a digital format help or hinder students with certain access-arrangement needs — see our companion piece on access arrangements), marking consistency, and resilience against technical failure at scale, with hundreds of thousands of students sitting the same exam on the same morning. That's plausibly why every announced timeline in this piece, across three very different systems, has moved rather than held.
Does going digital change what's actually being tested?
This is a genuinely open question rather than a settled one, and it's worth raising even though this article can't fully answer it. On one hand, every system covered here frames digitisation as a delivery change, not a content change — the intent is to test the same knowledge and skills, just on a screen instead of paper. On the other hand, Estonia's own reporting notes that its digital format can measure listening comprehension in a way a paper exam structurally cannot, which is a capability change, not just a format change, even if it wasn't the primary motivation. A parent weighing what "on-screen GCSEs" will mean for their own child should keep those two claims separate: a board saying "we're moving delivery on-screen" is not the same promise as "the exam will test the same things in the same way", and the honest answer, at least in the UK's case, is that we won't fully know until a pilot subject has actually run and been evaluated.
Comparison: where four systems stand on digital exams
| System | Status as of 2026 | Scope |
|---|---|---|
| England (AQA, Pearson Edexcel, OCR) | Piloting specific subjects/components, subject to Ofqual approval | A handful of named pilot subjects; full-scale rollout not expected before ~2030 |
| Estonia | Well advanced, staged rollout | Language exams moved first; some subjects (e.g. maths) still paper-based as of 2025 |
| Netherlands | Piloting via CvTE, further ahead on vocational (VMBO) tracks | Centrally set centraal examen component piloted first; school-run component unaffected |
| Singapore | Largely paper-based for national exams, with digital literacy assessed separately | No announced timeline for wholesale on-screen national exams as of 2026 |
Read across the table, the clearest pattern is that scale, not ambition, sets the pace. Estonia's population is under 1.4 million; the Netherlands administers a national exam system for a country of roughly 18 million but keeps half of every grade at school level; England, Wales and Northern Ireland between them run GCSEs for millions of students a year through several independent boards that each need Ofqual's separate sign-off before changing format. A smaller, more centrally coordinated system simply has fewer parties who need to agree before a pilot can run — which is a governance observation, not a comment on which country's students are better served by the pace they've chosen.
A parent's practical takeaway
If your child is currently in secondary school, the realistic expectation for the next few years is: mostly paper, with a small and growing number of on-screen pilot options in specific subjects, expanding cautiously and subject to regulatory sign-off at each step. Any specific claim about "GCSEs going digital in [year]" for a broad set of subjects should be checked directly against the relevant exam board's current published plans — this is a fast-moving and frequently revised area, and what's accurate in 2026 may already be out of date by the time you're reading this.