Can AI Fix Thailand's Math Crisis? OECD Study Maps Skills Gaps and Reform Priorities Ahead

An OECD study using AI-enhanced analysis finds that Thailand's mathematics challenges stem less from basic calculations and more from weaknesses in higher-order reasoning, algebra and real-world problem-solving, despite strengths in geometry and estimation. The report urges policymakers, development partners and businesses to use detailed learning diagnostics to drive targeted curriculum reforms, reduce socio-economic learning gaps and build a future-ready workforce for Thailand's knowledge-based economy.

Thailand's education challenge is no longer about getting children into classrooms, it is about ensuring they leave school with the skills needed for a modern, technology-driven economy. A new study by the Organisation for Economic Co-operation and Development (OECD) through its Centre for Educational Research and Innovation (CERI) offers a deeper look into this challenge by using artificial intelligence (AI) to analyse Thailand's performance in the 2022 Programme for International Student Assessment (PISA) mathematics assessment. Instead of relying solely on the country's average PISA score, the study examines how students perform across 17 cognitive skills and 12 mathematical content areas, providing policymakers with a far more detailed understanding of learning gaps. Based on responses from 8,495 students in 279 schools, representing about 604,600 Thai 15-year-olds, the report provides valuable insights for governments, development partners and private-sector stakeholders seeking to strengthen Thailand's human capital.

Looking Beyond Thailand's Overall PISA Score

Thailand recorded an average mathematics score of 393.9 in PISA 2022, well below the OECD average of 472.4. However, the report argues that average scores alone do not explain why students struggle. Using the OECD's Collective Intelligence Model for Education (CIME), AI and educational experts analysed every student response to identify strengths and weaknesses in individual mathematical skills.

The findings show that Thai students perform relatively well when solving familiar mathematical problems. They are comfortable applying known procedures, choosing suitable solution strategies and extending existing mathematical methods. However, they find it much harder to solve unfamiliar problems that require higher-order thinking. Students struggle to convert real-life situations into mathematical models, recognise relationships between variables, identify mathematical patterns and explain what their mathematical answers mean in practical situations.

The report also reveals differences across mathematics subjects. Students perform best in geometry, measurement, estimation, and probability, but weaker in algebra, equations, functions, arithmetic operations, numbers, units, and data interpretation. Since these topics form the foundation of science, engineering, digital technologies and artificial intelligence, strengthening these areas will be important for Thailand's future workforce.

Why These Findings Matter for Policymakers

The study suggests that Thailand's next phase of education reform should focus less on memorisation and more on developing analytical thinking, reasoning and problem-solving skills. The report shows that students can often perform calculations correctly once a problem has been presented mathematically, but many struggle to identify how mathematics applies to real-world situations.

For policymakers, this means curriculum reforms should encourage practical problem-solving, critical thinking and mathematical reasoning from an early stage. Teacher training programmes may also need to shift towards helping educators develop inquiry-based learning rather than relying mainly on procedural teaching methods.

Another major concern highlighted in the report is educational inequality. Students from higher economic, social and cultural status households consistently outperform disadvantaged students across almost every mathematical skill measured. In several competencies, achievement gaps exceed 80 score points, showing that socio-economic background remains one of the strongest predictors of educational success. Reducing these disparities through targeted support, improved school resources and stronger learning programmes could significantly improve national learning outcomes.

A New Opportunity for Development Partners and Investors

The report also highlights how artificial intelligence could become an important tool for education policy. Instead of replacing teachers, AI allows governments to analyse millions of assessment responses quickly while experts validate the findings. This creates detailed evidence that can guide curriculum improvements, teacher development and learning support programmes.

For international development partners such as the World Bank, Asian Development Bank, UNICEF, UNESCO and bilateral donors, the study provides a practical model for improving education investments. Rather than focusing only on school enrolment or infrastructure, development programmes can use AI-powered learning diagnostics to identify specific competency gaps, monitor progress and design targeted interventions that improve learning quality.

The findings are equally important for the private sector. Thailand's manufacturing, electronics, automotive, financial services and digital industries increasingly require workers with strong analytical skills, data literacy and problem-solving abilities. Weaknesses in algebra, mathematical reasoning and data interpretation could eventually contribute to skills shortages, affecting productivity and competitiveness. Businesses therefore have an opportunity to work more closely with schools and universities through internship programmes, curriculum partnerships, STEM initiatives and workforce development projects that prepare students for future employment.

Turning Assessment Data into Smarter Education Policy

One of the report's biggest contributions is demonstrating that international assessments like PISA can be used for much more than ranking countries. AI-powered analysis enables governments to identify exactly which mathematical skills require attention, making education reforms more targeted and evidence-based.

For Thailand, the study suggests that future reforms should strengthen algebraic thinking, mathematical modelling, data interpretation and higher-order reasoning while preserving existing strengths in geometry and spatial reasoning. Investments in teacher capacity, competency-based learning and AI-supported assessment systems could help ensure that students develop the skills needed for innovation-driven economic growth.

Beyond Thailand, the report offers an important lesson for many developing countries. As economies become increasingly digital and knowledge-based, education systems must move beyond measuring whether students know the correct answer and focus on whether they can analyse problems, apply knowledge and think critically. By combining artificial intelligence with expert educational analysis, governments and development partners have an opportunity to transform assessment data into practical policy solutions that strengthen human capital, improve economic competitiveness and prepare future generations for rapidly changing labour markets.