Articles in Original Research & Data
1 article
AI Buddy Insights is where we publish data for leaders of Cambridge and Pearson Edexcel schools. It covers our analysis of exam boards’ own results data and the public datasets worth knowing, and in time reports from AI Buddy and surveys of school leaders. Every figure is sourced, and every report states its method and its limits.
This is the hub for our Original Research & Data pillar. The first report, Cambridge June 2026 Results, is published. The others are listed below as they appear.
Why does original data matter to school leaders?
Because most decisions rest on comparisons: your results against the world, your market against competitors, your AI practice against peers. Boards and governments publish much of this data, but it is spread across press releases, PDFs and surveys with different methods. Our reports put it in one place, with the caveats that come with each source.
What did the June 2026 Cambridge results show?
Growth, led by A Level and by Asia. From Cambridge’s releases and Tes:
| Measure (June 2026) | Figure |
|---|---|
| Students sitting Cambridge exams | Nearly 705,000, in 148 countries |
| Total entries | 1.7 million |
| AS & A Level entries | About 737,000, up 5% |
| IGCSE entries | Almost 757,000, up 3% |
| O Level entries | Almost 231,000, up 2% |
| Fastest AS & A Level growth named | Indonesia 12%, China 7% |
| Students graded without exams (Gulf) | Almost 30,000 |
The full analysis, including world grade distributions for key subjects and what they mean for schools, is in Cambridge June 2026 Results.
Which public datasets should leaders know?
Five sources cover most questions leaders ask:
| Dataset | Publisher | What it tells you | Caveat |
|---|---|---|---|
| Results statistics (cumulative world totals) | Cambridge International, each series | Percentage of candidates at each grade, per syllabus | Cohort mix changes year to year; no all-subject total |
| Results releases | Cambridge International | Entries, growth, popular subjects, country highlights | Country figures are selective |
| Global International Schools Snapshot | ISC Research (commercial) | Market size: 15,075 schools, 7.7 million students, US$69.3 billion fee income (2026) | Proprietary definitions |
| TALIS 2024 | OECD | Teachers’ working time and AI use across 55 education systems | Lower secondary, mostly national systems |
| Technology in schools survey 2024–25 | UK Department for Education | Technology and AI use in schools in England | England only |
Market figures are from ICEF Monitor, reporting ISC Research. OECD’s TALIS 2024 found that “around one in three teachers report having used artificial intelligence (AI) in their work, on average across OECD education systems” (OECD). The DfE’s survey found that “less than half of teachers reported using generative AI (GenAI) for school activities (44%)” (DfE).
How should a school read Cambridge’s results statistics?
As a reference point, not a target. Cambridge publishes, for each syllabus, the cumulative percentage of candidates at each grade worldwide. For example, in June 2026, 16.9% of A Level Mathematics candidates achieved A* and 74.3% achieved A* to C (Cambridge). Three rules make the comparison fair:
- Compare like with like. Match syllabus codes, and for IGCSE, compare Core and Extended entry patterns.
- Allow for cohort size. A department of 15 moves several points with one student.
- Look at trends. Cambridge warns that “these figures may differ from previous June series” as the mix of schools and countries changes.
The method is in benchmarking school results against Cambridge worldwide statistics.
Which of its own numbers should a school track?
Public data only helps if the school’s own data is comparable. These are the internal measures we recommend tracking each series, as our guidance rather than any board’s requirement:
| Measure | Why it matters | Compare with |
|---|---|---|
| Grade distribution by syllabus code | The only like-for-like comparison with Cambridge’s world totals | Cambridge results statistics for the same series |
| Forecast accuracy (predicted vs achieved) | Tests the evidence behind UCAS predictions and forecast grades | The school’s own previous series |
| Tier entry patterns (Core/Extended, Foundation/Higher) | Entry decisions cap the grades available | Department targets and prior attainment |
| Entries per subject over three years | Shows demand shifts before they hit timetables | Board entry trends, e.g. Global Perspectives and IPQ growth |
| Topic-level mock performance | Explains where the grade gaps come from | Examiner reports for the syllabus |
| Resits and enquiries about results | Signals marking or entry problems | Board deadlines and outcomes |
Most schools already hold these numbers in spreadsheets or their management information system. The work is to keep them consistent from one series to the next, so that a change reflects students, not a change in how the data was collected. For methods, see question-level analysis of mocks and the predicted grades framework.
What do we know about AI in international schools?
Less than most headlines suggest. The best evidence comes from large national surveys, not international-school samples:
| Survey | Finding |
|---|---|
| OECD TALIS 2024 | About one in three teachers use AI at work, on average; “around 75% of teachers in Singapore and the United Arab Emirates” |
| OECD TALIS 2024 | ”Three in four teachers report that they lack the knowledge or skills to teach using AI” |
| DfE Technology in schools 2024–25 (England) | 44% of teachers used generative AI for school activities; 35% for lesson planning |
| UNESCO (2023) | “Less than 10% of schools and universities have formal guidance on AI”, from a survey of about 450 institutions |
For how schools are responding, see AI in Cambridge & Edexcel Schools.
What reports are in preparation?
Three, each with a method statement before any figure is published:
| Report | Data source | Status |
|---|---|---|
| The State of IGCSE Learning Gaps | Anonymised, aggregated AI Buddy question-level results across schools | In preparation; requires a privacy-reviewed data export |
| Most commonly missed topics, by syllabus (starting with Mathematics 0580) | Aggregated AI Buddy results by syllabus topic | In preparation; follows the learning gaps report |
| International School Leader Survey 2027 | Survey of school leaders on AI, workload and exam readiness | Planned with partner schools, January to March 2027 |
We will not publish AI Buddy figures until the method, sample and anonymisation are documented alongside them.
What standards do our reports follow?
Four commitments apply to every report in this pillar:
- Primary sources first. Exam boards, governments and regulators ahead of commentary.
- Method and sample stated. Who, how many, when, and what was excluded.
- No invented precision. Where a source rounds or a figure is uncertain, we say so.
- Privacy by design. Platform data is aggregated and anonymised; no school or student is identifiable.
How can schools use this data?
For three decisions in particular:
- Target setting. Compare predicted distributions with world statistics for each syllabus before results day.
- Curriculum planning. Growing subjects, such as Global Perspectives (IGCSE and O Level entries up 9% in June 2026) and the Project Qualification (up 68%), signal where family demand is moving.
- Market strategy. Entry growth by country shows where competition for students and teachers is rising. See country guides.
How schools do this with AI Buddy
The data that matters most is a school’s own. AI Buddy records mock and practice results by subject, topic and student, across Cambridge and Pearson Edexcel. Leaders can compare cohorts with world statistics and see which topics drive gaps. Aggregated, anonymised and with consent, the same kind of data will underpin our learning gaps reports.
Frequently asked questions
Does Cambridge publish grade distributions?
Yes. For each series Cambridge publishes cumulative world totals by syllabus, showing the percentage of candidates at each grade. There is no single all-subject figure.
How many international schools are there?
ISC Research counts 15,075 international schools worldwide in 2026, enrolling 7.7 million students, as reported by ICEF Monitor.
How many teachers use AI?
Around one in three lower-secondary teachers across OECD systems, according to TALIS 2024. In England, 44% of teachers used generative AI for school activities in the DfE’s 2024–25 survey.
When will AI Buddy’s learning gaps report be published?
It is in preparation. It will be published with its method, sample and anonymisation approach once a privacy-reviewed data export is complete.
Discover how AI Buddy helps schools strengthen teaching, learning and evidence-informed school improvement. Or start a short consultation with our schools team using the form below — we will get back to you directly.
Sources
- Cambridge International, AS & A Level results release (11 August 2026), IGCSE and O Level results release (18 August 2026) and A Level results statistics June 2026
- Tes, Exam results 2026: Cambridge International A levels released (11 August 2026)
- ICEF Monitor, Continuing expansion of K-12 international school sector (February 2026), reporting ISC Research
- OECD, Results from TALIS 2024: Teaching for today’s world (October 2025)
- Department for Education, Technology in schools survey 2024 to 2025 (November 2025)
- UNESCO, UNESCO survey: less than 10% of schools and universities have formal guidance on AI (June 2023)