Türkiye’s Artificial Intelligence Action Plan for 2026–2030 took effect after being published in the Official Gazette under a presidential circular, turning the goals announced in June into a formal framework binding on public institutions and agencies. The plan was prepared under the coordination of the Ministry of Industry and Technology, with the participation of the public and private sectors, universities, and civil society organizations.
All public institutions and entities are required to implement the responsibilities set out in the plan and provide the necessary support during the implementation phase. The plan expects the public and private resources allocated to it to generate economic value exceeding 1 trillion Turkish lira.
Four Pillars and 16 Priority Actions
The plan is based on four pillars: “awareness,” “utilization,” “production,” and “governance,” with each pillar comprising four actions. The first pillar focuses on artificial intelligence literacy, human resources, the data ecosystem, and the legal framework. The utilization pillar addresses building computing infrastructure and expanding the use of artificial intelligence in the public and private sectors.
The production pillar covers financing, local models, robotics, and artificial intelligence growth zones, while the governance pillar addresses international investments, artificial intelligence-related diplomacy, regional cooperation, and security controls.
Expanding the Skills Base and Infrastructure
The plan aims to establish an artificial intelligence literacy assessment system integrated with the e-Government platform “e-Devlet,” enabling citizens to measure their levels and find suitable training programs. A total of 1.5 million people are expected to complete the national artificial intelligence literacy program during the first 12 months, with the number reaching 5 million within two years.
The plan also includes organizing workshops in Türkiye’s 81 provinces and training at least 1,000 instructors during the first six months. By the end of 2027, Türkiye aims to prepare 10,000 advanced specialists and 100,000 practitioners for artificial intelligence applications, and to include core artificial intelligence courses at no fewer than 50 universities.
In the field of data, the plan aims to publish at least 2,000 government datasets in machine-readable format in the National Data Library, with at least 100 of them made regularly available through application programming interfaces.
What Will Change in Practice?
The plan places computing capacity at the center of implementation. It aims to raise the combined capacity of data centers and artificial intelligence infrastructure to 1 gigawatt by 2030, provide at least 1 exabyte of storage capacity, connect the national backbone at a speed of 100 terabits per second, and establish low-carbon electricity purchase agreements totaling 10 terawatt-hours annually.
It also aims to stimulate at least $10 billion in private investment in data centers, cloud services, and artificial intelligence infrastructure. The “GPU for Everyone” program will provide computing credits to startups, small and medium-sized enterprises, and researchers, starting with at least 2 million GPU hours annually and rising to 20 million hours annually by the end of 2028.
Corporate Financing and Sectoral Models
The plan calls for the establishment of a two-stage financing structure. The National Artificial Intelligence Research Fund aims to provide at least 10 billion Turkish lira, while the Artificial Intelligence Growth Fund targets 15 billion lira and participation in financing at least 20 startups in Series A and Series B rounds or expansion stages.
By the end of 2028, Türkiye aims to announce at least five artificial intelligence growth zones and operate three of them, in addition to establishing three regional centers of excellence and six access points. A total of 1,000 artificial intelligence vouchers will also be allocated to small and medium-sized enterprises during the first 12 months.
The targets include developing foundation models for the health, agriculture, earthquake, energy, and industrial sectors, along with applications for robotics, autonomous driving, manufacturing, agriculture, healthcare, and disaster technologies. Türkiye also aims to develop a shared large language model for Turkic languages covering the Oghuz, Kipchak, and Karluk families, and launch its first version in a production environment by the end of 2027.
Regulatory Sandboxes and Public Monitoring
Regulatory sandbox environments will be established in at least five priority sectors, including finance, health, energy, mobility, and communications. High-impact artificial intelligence systems will be subject to algorithmic impact assessments, model cards, and safety testing.
The National Artificial Intelligence Council will oversee implementation, with institutional performance in terms of targets, budgets, and timelines published through a public progress portal. The first report is scheduled to be issued within 12 months of the plan’s publication, followed by annual reports.