Is Computer Science Still Worth It With AI? A Teacher's View
Is GCSE or A Level Computer Science still worth taking with AI? A teacher's balanced look at 2026 entry trends, UK jobs evidence and university requirements.
By Yuness, qualified Computer Science teacher and former Head of KS5 Computer Science · Last reviewed 21 November 2026
In my view as a Computer Science teacher, yes. For most students who enjoy solving problems and are comfortable with maths, Computer Science is still worth taking at GCSE, A Level or IB, and AI makes understanding how computers, data and algorithms work more useful, not less. The honest caveat is that the easy route ("learn a bit of coding, get a coding job") is narrowing: entry-level tech hiring has become tougher, and AI is changing what junior developers do. If you choose it, take it for the thinking it teaches, not as a guaranteed job ticket.
| Question | Short answer (as of November 2026) |
|---|---|
| Is it still worth taking? | Yes, if you like logic, puzzles and maths and want to understand the technology you use every day. |
| Are fewer students taking it? | At GCSE, yes: entries in England fell 7% in 2026. At A Level, entries rose slightly. |
| Is the job market harder? | At entry level, yes. But demand for AI, data and cyber security skills is still growing. |
| Will AI do the coding for me? | It already helps with well-defined tasks. Deciding what to build, checking the output and handling messy real problems is still human work. |
| Do I need it for a Computer Science degree? | At Oxford, Cambridge and Manchester, Maths is required and A Level Computer Science is not, though many successful applicants take it. |
What is happening to GCSE and A Level Computer Science entries?
The headline story of 2026 is a fall at GCSE and a small rise at A Level. Ofqual's provisional entry statistics for England (published 9 June 2026) show:
| Qualification (England) | Summer 2025 | Summer 2026 | Change |
|---|---|---|---|
| GCSE Computing | 89,610 | 83,330 | down 7.0% |
| A Level Computing | 19,010 | 19,685 | up 3.5% |
In the provisional data, computing had the largest percentage fall of any EBacc subject at GCSE. Schools Week noted that GCSE entries peaked at 93,980 in 2024, so more than 10,000 fewer students are now taking it than two years ago. The final results told the same story: JCQ figures reported by Tes show GCSE computing entries across England, Wales and Northern Ireland down 7.1%, from 91,619 in 2025 to 85,096 in 2026. At A Level, BCS analysis of JCQ data shows Computer Science entries in England up 2% in 2026.
Is AI putting students off?
Possibly, but nobody knows for sure. Speaking to Tes on 21 August 2026, Myles McGinley, managing director of OCR, said he was speculating, but that AI may be leading to "students thinking they don't need to learn how to code". He also argued the subject is more important than ever in the age of AI. The National Centre for Computing Education's GCSE results analysis lists AI debates as only one possible factor, alongside teacher recruitment and retention and the pandemic disruption this cohort experienced. Entries are also still above their 2019 level.
What does the UK job market evidence actually say?
This is where honesty matters most. The evidence points in two directions at once, and both are true.
Entry-level roles have become harder to get
- The Institute of Student Employers reported in October 2025 that graduate hiring fell 8% year on year, while apprentice hiring rose 8%. Its latest survey, reported by Times Higher Education in October 2026, found graduate hiring recovered by 4% in 2025-26, but competition rose to 141 applications per graduate role.
- A DSIT assessment of AI and the UK labour market (January 2026) notes that UK digital sector employment fell in 2024 for the first time in a decade, and the number of 16 to 24-year-olds in computer programming fell 44% in a single year. The same report stresses that these patterns are suggestive but do not yet show that AI is the cause.
- In the US, Stanford's Digital Economy Lab reported in August 2026 that employment for 22 to 25-year-olds in the most AI-exposed occupations was about 19% below where it would otherwise be, mainly through reduced hiring rather than layoffs. The authors found no widespread displacement overall and describe their findings as descriptive, not proof that AI caused the gap.
Demand for computing skills has not gone away
- Indeed's mid-2026 UK hiring report found 9.4% of UK job postings mentioned AI by the end of June 2026, a record, with the highest shares in data and analytics and in software development. IT systems and solutions had the fastest posted wage growth, at 7.2% year on year in the three months to June. This is against a weak overall market: total UK postings were 32% below their February 2020 level.
- The government's cyber security skills report (September 2026) found core cyber security job postings rose 7% in 2025 compared with 2024, a partial recovery after a 33% fall in 2024.
- DSIT's AI Labour Market Survey 2025 found 35% of organisations struggling to fill AI roles. The biggest skills gap was understanding AI concepts and algorithms, and the report notes that computer science remains the dominant qualification for AI roles.
- Government-commissioned projections by the Warwick Institute for Employment Research (January 2026) put programmers and software development professionals at the top of the list of occupations expected to integrate AI into their work by 2035, around 555,000 jobs in one scenario. This is a projection of jobs that will involve AI, not a forecast that every one will be a new job.
My reading as a teacher: AI is not ending computing careers, but it is raising the bar. Employers are looking for people who understand systems, data and AI properly, not just people who can write a few lines of code. That is a reason to study the subject seriously, not a reason to avoid it. The evidence is still early and mixed, though, so treat this as a judgement rather than a prediction.
What about university applications?
Fewer students are going on to computing degrees. BCS analysis of UCAS data shows undergraduate computing acceptances fell 7% in 2026, while acceptances across all subjects rose 2%. Within computing, interest is shifting: BCS reported in April 2025 that applications for AI degrees rose by 15% among UK women and 12% among men, while computer science and software engineering applications declined.
What AI is changing, and what it is not
The DSIT assessment gives a useful, measured picture:
- Changing: the length of coding tasks AI agents can complete on their own has roughly doubled every seven months. In one evaluation of over 1,300 professional tasks, the best model was rated as good as or better than an expert in nearly half of cases. Studies of software development report large time savings.
- Not (yet) changing: those evaluations used precisely specified, self-contained tasks. The report notes that real work involves more ambiguity and iteration, and that results from experiments may not carry over to real workplaces.
My interpretation as a teacher: AI is good at the part of programming that was already well defined. Working out what the problem actually is, breaking it down, choosing a sensible design, testing it and spotting when an AI-generated answer is subtly wrong all depend on understanding. That understanding is exactly what a good Computer Science course builds. The Stanford team found the same pattern from another angle: where AI complements workers rather than automating their tasks, employment has been flat or rising.
What you actually learn in Computer Science (it is not just coding)
Many people imagine Computer Science is a coding course. It is much broader. The OCR GCSE (J277) specification has two 1 hour 30 minute written papers:
- Computer systems: systems architecture, memory and storage, networks, network security, systems software, and the ethical, legal, cultural and environmental impacts of technology.
- Computational thinking, algorithms and programming: algorithms, programming fundamentals, producing robust programs, Boolean logic, and programming languages and IDEs.
The OCR A Level (H446) specification goes further: processors and storage, software development, exchanging data, data types, data structures and algorithms, computational thinking, problem solving and programming, plus legal, moral, cultural and ethical issues. Two written papers make up 80% of the grade and a programming project the other 20%. OCR states that every A Level Computer Science qualification must contain at least 10% mathematical skills.
The skills that carry over into almost any career are problem decomposition, abstraction (deciding what detail matters), logical reasoning, precision, and understanding how systems fit together. You can explore every topic on our free GCSE Computer Science hub and OCR A Level Computer Science hub.
How AI tools fit into studying Computer Science
AI cannot sit the exam for you. Both J277 papers are written, non-calculator exams, and OCR assesses practical programming skills in Paper 2, particularly Section B. At A Level, the two written papers are 2 hours 30 minutes each. So the grade depends on what is in your head.
Used well, AI can be a patient study partner. My advice to students:
- Write your own attempt first, then ask an AI tool to explain where it went wrong, rather than asking it to write the code.
- Ask it to explain code line by line, then close it and rewrite the program from memory.
- Treat its answers as claims to check: run the code, test edge cases, and compare with your notes.
- Follow your school's and exam board's rules on AI for any coursework.
AI is also becoming part of what you study. The new IB Computer Science course, first examined in May 2027, includes a topic on machine learning (our guide to machine learning fundamentals covers it) (see our IB Computer Science hub). In England, the Curriculum and Assessment Review recommended that a broader GCSE Computing replace GCSE Computer Science and include AI. As of OCR's January 2026 briefing, no firm start date had been given. Our guide to the new Computing GCSE explains the changes in more detail.
Who should take Computer Science, and who might choose differently?
It is a great choice if you:
- enjoy puzzles, logic and working out why something does not work;
- are reasonably confident in maths, or willing to work at it;
- want to understand the technology and AI you use, not just use it;
- are considering computing, engineering, maths, data science, cyber security or a digital apprenticeship;
- like the idea of building something and improving it step by step.
You might choose differently if you:
- mainly want to use software creatively (design, media, music) rather than understand how it works;
- strongly dislike maths and precise, logical work;
- are choosing it only because "tech pays well", without enjoying the problem solving;
- (Year 11) are aiming for a Computer Science degree but would have to drop A Level Maths to fit it in. Maths matters more for most courses.
For parents: your child does not need to have coded before. OCR's GCSE specification states that no prior knowledge of the subject is required. What helps most is a willingness to keep going when a program does not work, and that habit is valuable whatever they do next.
What do universities ask for? Maths vs Computer Science
For Computer Science degrees, Maths is usually the key subject. Three examples, checked on official pages in November 2026:
| University | Typical A Level offer | Required subjects | Is A Level Computer Science required? |
|---|---|---|---|
| Oxford | A*AA, including A*A in Maths and Further Maths if available | Maths required; Further Maths highly recommended | No |
| Cambridge (2027 entry) | A*A*A | Maths, and Further Maths where your school offers it | No, but 64% of recent entrants with A Levels had taken Computing |
| Manchester (BSc Computer Science, 2027 entry) | A*AA, with the A* in Maths | Maths, plus at least one of Computer Science, Further Maths, Biology, Chemistry or Physics | No: it is one of five subjects that can meet the science requirement |
Requirements differ between universities and change from year to year, so always check the official course page for your entry year before choosing A Levels.
FAQ
Will AI replace programmers?
The current evidence does not show that. DSIT's January 2026 assessment found AI is strong at precisely specified tasks but that real work involves more ambiguity, and it judged evidence of AI-driven job losses suggestive but not conclusive. Early-career hiring has become tougher, so my reading is that the job is changing: programmers increasingly direct, check and integrate AI-generated code, which needs more understanding, not less.
Do I need GCSE Computer Science to take A Level Computer Science?
Not for the qualification itself: OCR's A Level specification states there are no prior qualification requirements. However, sixth forms set their own entry criteria, and the GCSE gives you a head start with programming and theory, so check with your school. If you missed the GCSE, our GCSE hub is a good place to catch up.
Is Computer Science or Maths better for a Computer Science degree?
Maths. Oxford, Cambridge and Manchester all require A Level Maths, and none of them requires A Level Computer Science. Oxford highly recommends Further Maths, and Cambridge requires it (to AS or A Level) where your school offers it. Computer Science is still a useful third subject: at Cambridge, 64% of recent entrants with A Levels had taken it.
Is GCSE Computer Science being replaced?
The Curriculum and Assessment Review recommended a new, broader GCSE Computing to replace it, including AI. As of January 2026 no firm date had been set. See our explainer on the new Computing GCSE for the latest.
Is GCSE Computer Science hard?
It is demanding but achievable. In England in 2026, 70.6% of students achieved grade 4 or above and 29.9% achieved grade 7 or above, according to the National Centre for Computing Education. The challenge is usually regular programming practice and precise exam answers, not natural talent.
If you are weighing up your options, explore the free GCSE, A Level and IB Computer Science hubs to get a feel for the subject. If you would like to talk it through or try some real Computer Science with a qualified teacher, you can book a free 1-hour trial lesson.