Free access to the best AI tools for students and teachers, ensuring that the benefits of technology reach everyone, not just schools with better resources.
Measuring Effectiveness
What do we measure?
The goal of the AI Leap educational innovation program is to support the development of students’ thinking and learning skills through meaningful use of artificial intelligence. To ensure that AI empowers rather than hinders learning, and that it does so in a way that fits the Estonian language and cultural context.
AI Leap has three main objectives:
Equality and accessibility
Development of learning skills
Creation of tutor-style AI tools that support not only subject knowledge but also self-management, learning motivation, and persistence.
Advancement of AI literacy
Helping teachers and students become smarter and more responsible users of artificial intelligence.
Short-term indicators and results – monitored by the AI Leap Foundation
The effectiveness of AI Leap is evaluated regularly, using the goals of the 2025/26 school year as benchmarks. Progress is measured according to activities in February, April, and May 2026.
Students
| Goal for the 2025/26 school year | Target level (Apr 2026) | Results (Jan 2026) | Results (Feb 2026) |
|---|---|---|---|
| Technology accessibility | |||
| Free accounts created in the learning app for grade 10–11 students (20,000 users) | 75% | ||
| The app meets Estonian language and cultural guidelines | 70% | ||
| The app meets accessibility standards (complies with WCAG 2.1 accessibility criteria) | 80% | ||
| Support for AI adoption | |||
| Students have activated their user accounts | 80% | ||
| Repeated app usage (within 30 days) | 60% | ||
| Percentage of successful sessions | 60% | ||
| AI literacy level | |||
| Students’ positive AI learning experience | |||
| Student satisfaction with the learning app (on a scale of 1–5) | 4,2 | ||
| AI responses were clear and easy to follow (on a scale of 0–1) | 0,7 | ||
| Change in learning habits (self-assessment) | 4,0 | ||
| Improvement in learning skills | |||
| App compliance with evaluation criteria | 6,2 | ||
| Co-creation workshops held in schools (target: May 2026) | 60% | ||
| AI student programs implemented in schools (target: May 2026) | 70% | ||
Teachers
| Goal for the 2025/26 school year | Target level (Feb 2026) | Results (Oct 2025) | Results (Nov 2025) |
|---|---|---|---|
| Technology accessibility | |||
| The premium versions of leading language models have been made freely available to upper secondary schools (157 in total). | 75% | 99% | 99% |
| Free premium accounts have been created in the app for all grade 10–11 teachers (4,700 users). | 75% | 99% | 100% |
| The app meets accessibility standards (complies with WCAG 2.1 accessibility criteria). | 80% | - | - |
| Support for AI adoption | |||
| Teachers have activated their user accounts | 80% | 43% | 56% |
| Average daily number of unique users | - | 1771 | 1896 |
| Number of users in the AI Leap Moodle environment | 60% | 23% | 25% |
| Teachers perceive artificial intelligence as useful in their work | - | - | - |
| Supporting the rethinking of teaching | |||
| Number of school leaders who have participated in the foundation’s training programs | 80% | 90% | - |
| AI learning communities for teachers established in schools | 90% | - | 25% |
| Teachers participating in AI learning communities: professional learning community engagement | 80% | - | - |
| Readiness to use AI (knowledge, access, etc.) | - | - | - |
| Teacher satisfaction with AI use (on a scale of 1–5) | 4,2 | - | - |
| Teacher satisfaction with learning communities | 4,2 | - | - |
| Teacher satisfaction with materials provided by the foundation | - | - | - |
| Change in teaching methods (self-assessment) | 70% | - | - |
| Clear AI guidelines in schools | |||
| Number of schools that have created principles or guidelines for AI use | 100% | - | - |
Long-term evaluation
The long-term impact assessment of AI Leap is coordinated by the Ministry of Education and Research. The research is conducted by educational psychologists and researchers from the University of Tartu, led by Jaan Aru.
The study focuses on changes in students’ learning skills, including:
Autonomous motivation
Learning-related beliefs
Task persistence
Learning strategies
Learning-related self-regulation
Prior knowledge activation
Conceptual change
Learner growth in these areas is assessed directly from chat-logs, validated with real-world standardized testing, using a large-scale randomized controlled trial. Model behaviour is steered by student outcomes. The goal is to improve student behaviour, and research is conducted to find whether improvement in AI behaviour leads to improvement in student behavior.
