Azerbaijan's government opened a fully scholarship-funded AI Academy in Baku on 8 September 2025, with guest-faculty ties to MIT, Harvard, Stanford and Carnegie Mellon, unveiling a 10-billion-token Azerbaijani-language model prototype at launch.
10 billion tokens
Tokens used to train the prototype Azerbaijani-language model
1.2 million
Synthetic data points used to train the prototype model
Details
Promoter
AI Academy (Ministry of Digital Development and Transport / Ministry of Science and Education, supported by AZCON Holding)
Period
Launched 8 September 2025, ongoing
Keywords
government, higher education, AI research, digital skills
Context
The AI Academy opened in Baku on 8 September 2025, with the ceremony attended by Rashad Nabiyev, Minister of Digital Development and Transport, and Emin Amrullayev, Minister of Science and Education. It was established under Azerbaijan's Artificial Intelligence Strategy for 2025–2028 and operates with the support of AZCON Holding.
Activities
All educational programmes at the Academy are fully scholarship-funded. Curricula are described as developed in line with international standards, and the Academy has established relationships with MIT, the University of California, Carnegie Mellon University, Harvard University and Stanford University, whose faculty are invited as guest lecturers. At launch, the Academy presented a prototype Azerbaijani-language model trained on 10 billion tokens and 1.2 million synthetic data points, a concrete technical deliverable alongside its training mission.
Conclusions
Neither source publishes enrollment figures or independent evaluation of graduate or learning outcomes, so the Academy's actual training effectiveness remains unverified.
Implementation
Indicative cost
High (€500k–€5M)
Time to results
Long (> 3 years)
Staffing & skills
Ministry of Digital Development and Transport (Azerbaijan), Ministry of Science and Education (Azerbaijan), AZCON Holding (corporate sponsor), Guest faculty from MIT, University of California, Carnegie Mellon University, Harvard University, Stanford University
Conditions for success
backed by a national AI Strategy (2025–2028) with two ministerial sponsors
fully scholarship-funded programmes removing direct cost barriers to enrollment
partnerships with five leading global universities for guest lecturers
dedicated corporate sponsorship (AZCON Holding)
Common failure modes
no enrollment or graduate-outcome data has been published, so training effectiveness is unverified as of launch
Where it fits
Governance type
national government ministries with corporate sponsor
Scale
national training institute
Income level
upper-middle-income country (Azerbaijan)
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.