evidoria

← Back to browse

Good practice Imported

AI Integration and Curriculum Innovation — Azerbaijan's National Digital School Platform, Evaluated Across 25 Baku, Ganja & Sumgayit Schools

Azerbaijan · Baku · See the Azerbaijan profile · See the Baku profile

Evidence: Observational / pre–post Top 31% 60/100 · Ask Evidence Copilot about this practice

A 2026 peer-reviewed study of 1,250 students and 250 teachers across 25 schools in Baku, Ganja and Sumgayit found AI curriculum integration significantly predicted independent learning skills, alongside Azerbaijan's OpenAI-backed Digital School platform for 500,000+ students.

1,250
Students surveyed (grades 9-11)
250
Teachers surveyed
25
Schools surveyed
β=.55, p<.001
AI integration → curriculum innovation effect
β=.51, p<.001
Curriculum innovation → independent learning skills effect
B=0.15, p<.001
Teacher AI-literacy moderation effect
500,000+
Students targeted by Digital School platform
AI Integration and Curriculum Innovation — Azerbaijan's National Digital School Platform, Evaluated Across 25 Baku, Ganja & Sumgayit Schools

Details

Maturity
Scaling
Promoter
Azerbaijan Ministry of Science and Education; OpenAI; Baku State University; Shaanxi Normal University
Period
2025–2026
Keywords
K-12 education, EdTech, public policy

Context

Azerbaijan's Ministry of Science and Education, together with the Ministry of Digital Development and Transport, signed a memorandum of understanding with OpenAI to build an AI-based adaptive learning platform inside the national 'Digital School' project, aiming to give more than 500,000 students diagnostic assessment and individualised learning support. Separately, the Ministry partnered with G42's Presight and Microsoft Copilot Studio to add an AI teaching-assistant agent to the same platform.

Objectives

A July 2026 peer-reviewed study set out to test whether AI integration in schools predicts perceived curriculum innovation and, in turn, students' independent learning skills, and whether teachers' own AI literacy moderates that relationship.

Activities

The study surveyed 1,250 secondary students (grades 9-11) and 250 teachers across 25 public schools in Baku, Ganja and Sumgayit, using hierarchical regression and mediation analysis. The study deliberately sampled urban schools only, reasoning that they have greater technology access and investment.

Results

AI integration significantly predicted perceived curriculum innovation (β=.55, p<.001), which in turn predicted students' independent learning skills (β=.51, p<.001; indirect effect 95% CI 0.18-0.26). Teacher AI literacy significantly moderated this relationship (interaction B=0.15, p<.001), showing that teachers' own AI competence shapes how much benefit reaches students.

Conclusions

Because the study sampled only urban schools, its findings should not be generalised to rural Azerbaijan, and the cross-sectional, correlational design falls short of a randomised trial, so causal claims about learning impact remain limited.

Implementation

Indicative cost
High (€500k–€5M)
Time to results
Medium (1–3 years)
Staffing & skills
Classroom teachers with varying AI literacy, Baku State University and Shaanxi Normal University researchers

Conditions for success

  • Teacher AI literacy is a significant moderator of student benefit, implying teacher training is a precondition for impact
  • National ministry-level MOU with OpenAI and a separate G42/Microsoft Copilot Studio partnership providing platform infrastructure

Common failure modes

  • Study sampled urban schools only, deliberately excluding rural Azerbaijan from the evidence base
  • No independent audit of data governance or student-data protection was found for the platform

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful