Once again, maths is the most popular A-level subject. There were 115,380 entries in the subject in 2026, an increase of 3.9% from 2025. Further maths AS-level entries enjoyed an 8.5% increase from 2025 as well.
Moreover, recently published data from university admissions service Ucas for the 2026 application cycle showed a marked 11% increase in applications to study mathematical sciences at university.
All of these figures show that maths remains not just a hugely popular subject at school, but that more students are wanting to pursue it further.
In my opinion as a research mathematician, the staggering leaps that generative AI has made means there is no better time to be studying mathematics.
The possibilities of AI
AI is solving mathematics problems that have stood for decades. In May, OpenAI announced a counter-example to Paul Erdős’ planar unit distance problem, which stood unsolved for more than 80 years. In July, Anthropic’s LLM Claude Fable 5 found a counter-example to the Jacobian conjecture, which has stood since 1939.
Most recently, OpenAI announced “ten advances in mathematics and theoretical computer science” made with their internal model Astra. Very few would have predicted that AI would make progress on a vast range of difficult problems that have baffled mathematicians for decades.
This latest development has been met with enthusiasm and anxiety in equal measures in the mathematics community. Many of my colleagues are excited by the prospect that we’ll soon know the answers to many more long-standing problems. Others are worried that it spells the end of human-led mathematics research.
In their 2024 book AI Snake Oil: What Artificial Intelligence Can Do, What It Can’t and How to Tell the Difference, computer scientists Arvind Narayanan and Sayash Kapoor state that “AI is no threat to education, any more than the introduction of the calculator was”.
This sentiment reflects my feelings too. Just as the calculator with its superior levels of computation didn’t stop people from doing calculations, AI producing research-level maths will not stop people from doing, writing, creating and teaching this beautiful subject.
But AI should always remain a smart tool and not an artificial colleague, as philosopher Daniel Dennett used to say. If used correctly and prudently, AI can be a great enabler of potential and catalyst for progress in mathematics. Conjectures may now be resolved in days or weeks, not years or decades.
With AI, the field of mathematics can expand its horizons and find new applications to solve existing problems in areas such as data science, quantum computing, cybersecurity and the economy.
But perhaps most importantly, AI can lower the barrier to studying mathematics at university and make it easier to become a research mathematician. The generation of students entering university today will be able to learn, create, understand and ultimately do more with mathematics than any previous generation in history.
Terence Tao, recipient of the Fields medal, mathematics’ most prestigious prize, has led large-scale collaborative AI-assisted projects involving mathematicians at different stages of their careers. This is great reason for cautious optimism.
I say cautious, because these developments in AI necessitate a debate over what the future of mathematics will look like, both in research and education, which will naturally be informed by future progress in AI.
This debate will not be settled by the current generation of mathematicians alone. It will need the input and guidance of the young people who wish to enter the field and make careers in it.
Maths is less accessible
However, maths departments in the UK are threatened with cuts and closures. This leads to the creation of maths deserts: places in the country where there is little opportunity to study mathematics. This puts the subject out of reach for a swathe of the thousands of students who live at home while studying rather than moving to a university.
The Campaign for Mathematical Sciences, of which I am a supporter, maintains a provision tracker for higher education institutions facing pressure. Lower-tariff institutions are particularly vulnerable to cuts and closures, since the more prestigious higher-tariff institutions, mainly located in larger cities, have increased their intake of mathematics students especially since the pandemic.
The growing phenomenon of maths deserts has made the possibility of studying a maths degree in a non-metropolitan area even more difficult. This is particularly concerning, as graduates from lower-tariff universities are much more likely to go into teaching to educate the next generation of mathematicians.
The recent trends in A-level entries and university enrolments show that mathematics remains hugely popular. Universities should be capitalising on this enthusiasm, especially as new and exciting discoveries are on the horizon.
By Neil Saunders, Senior Lecturer in Mathematics, Department of Mathematical Sciences, City St George's, University of London. This article is republished from The Conversation under a Creative Commons license. Read the original article.
