Asian high-school students have once again emerged as global leaders in reading, mathematics and science, according to the 2025 Programme for International Student Assessment (PISA).
The latest assessment by the Organisation for Economic Co-operation and Development (OECD) also highlights a troubling trend: student performance in all three core subjects has declined globally. The results raise concerns about whether education systems are adequately preparing teenagers for an increasingly technology-driven economy.
Singapore topped the overall rankings, while students from China’s major economic hubs, including Beijing and Shanghai, also recorded some of the highest scores.
Which countries performed best in the 2025 PISA rankings?
The 2025 PISA results show Asian education systems continuing to dominate the top of the international rankings.
Singapore was among the strongest-performing participants across the three core subjects. Students from several major Chinese jurisdictions also performed exceptionally well.
The results are particularly notable because PISA measures the ability of 15-year-olds to apply knowledge to real-world problems rather than simply testing whether students can recall information.
The leading Asian systems have therefore become important benchmarks for countries seeking to improve mathematics, science and reading outcomes.
Why are US reading scores raising concerns?
The United States recorded a particularly worrying result in reading.
American secondary students’ reading performance fell to levels approaching a 25-year low, according to the information provided. Their scores also left them behind students from the four Chinese jurisdictions that participated in the assessment.
The decline comes at a strategically important moment for both countries.
The US and China are competing for leadership in areas such as artificial intelligence, advanced manufacturing and other emerging technologies. Both countries reportedly spent more than $1 trillion on research and development in 2024.
A weaker educational foundation could eventually affect the supply of skilled workers needed to compete in these industries.
How is artificial intelligence changing education?
The PISA report also raises a larger question about the role of artificial intelligence in classrooms.
AI tools can give students instant access to explanations, summaries and answers. But the OECD has warned that access to more information does not automatically translate into better learning.
Andreas Schleicher, the OECD’s director for education and skills, pointed to concerns about declining critical-thinking abilities.
The challenge is particularly relevant in an environment where students can use AI to generate information within seconds. Students still need to determine whether information is accurate, identify relevant evidence and distinguish useful knowledge from misinformation.
Why critical thinking matters in the AI era
Traditional education often focused on acquiring a defined body of knowledge. AI changes the equation because students can increasingly obtain information without memorising it themselves.
That makes other skills more important, including:
- Evaluating whether information is reliable.
- Identifying misleading or incomplete claims.
- Comparing competing sources.
- Asking better questions.
- Applying information to unfamiliar problems.
- Explaining why an answer is correct rather than simply producing one.
The concern raised by the OECD is therefore not necessarily that students have access to technology. It is whether technology is strengthening or weakening their ability to think independently.
Are digital devices hurting classroom performance?
PISA also examined the effect of digital distraction during lessons.
The findings point to a notable difference between some high-performing Asian education systems and the global average.
Fewer than 10% of students in China, Japan and South Korea reportedly said their peers were distracted by digital devices during science lessons. The global average was about 28%.
That gap suggests that classroom technology policies and student behaviour may be relevant to academic performance.
However, the figures should not be interpreted as proof that restricting smartphones alone will improve test scores. Educational outcomes depend on many factors, including teaching quality, curriculum, family circumstances, student motivation and school resources.
Is family wealth still the biggest factor in student performance?
The report also suggests that the relationship between student performance and national or family wealth is becoming weaker as technology spreads.
Digital access can give students information and learning resources that were previously available mainly to wealthier households or better-funded schools.
But access alone does not guarantee better outcomes.
A student with a laptop and internet connection may have enormous amounts of information available but still struggle to evaluate it, concentrate during lessons or develop strong foundational skills.
The emerging challenge for education systems is therefore moving from simply providing technology to teaching students how to use it effectively.
What could declining student performance mean for the economy?
Education outcomes have consequences far beyond schools.
A Stanford University study cited in the supplied information estimated that bringing all 15-year-old Americans up to basic PISA standards could add more than $40 trillion to the US economy over their working lives.
The estimate illustrates the enormous economic stakes associated with educational performance.
Better foundational skills can translate into a more productive workforce, stronger innovation and greater capacity to adapt to technological change.
At the same time, the PISA data point to a worrying deterioration across OECD countries. The share of OECD students performing below the baseline level reportedly rose from 16% in 2022 to 20% in 2025.
That means the issue is not simply about which country ranks first. A larger concern is whether education systems worldwide are producing enough students with the skills required for modern economies.
Why are Asian education systems outperforming Western counterparts?
The PISA rankings should not be reduced to a simple East-versus-West comparison.
High-performing education systems differ in their curricula, teacher training, examination structures, family expectations and approaches to classroom discipline.
Singapore, China, Japan and South Korea have each developed education systems with strong emphasis on foundational academic skills, although their approaches are not identical.
The results also do not mean that students in these systems are universally better educated in every respect. PISA measures specific competencies at a particular age and cannot capture every dimension of education, including creativity, wellbeing, vocational skills or social development.
Nevertheless, the continued presence of Asian systems near the top of international assessments provides useful evidence for policymakers examining what works in mathematics, reading and science education.
What does the PISA data mean for the future of education?
The 2025 results point to two problems developing at the same time.
First, basic academic performance is weakening in many education systems. Second, students are entering an era in which they will need stronger critical-thinking skills because AI can increasingly handle information retrieval and routine cognitive tasks.
That combination makes the next phase of education particularly important.
Schools may need to place greater emphasis on foundational literacy and numeracy while also teaching students how to work with AI without becoming dependent on it.
The most important lesson from the PISA results may therefore be less about which country sits at the top of the table and more about what students are being prepared to do once they leave school.
In an economy increasingly shaped by artificial intelligence, simply knowing more information may not be enough. Students will need to know how to question it, test it and use it.
