Türkiye AI Action Plan 2030

Overview

Türkiye Yapay Zekâ Eylem Planı (2026-2030), put into effect through Presidential Circular No. 2026/9 dated 18 August 2026, is Türkiye’s national policy roadmap for artificial intelligence for the 2026-2030 period. The Circular states that the Plan was prepared under the coordination of the Ministry of Industry and Technology with the cooperation of public bodies, the private sector, universities and civil-society organisations. It directs all public institutions and organisations to perform their responsibilities under the Plan diligently and to provide the support needed for implementation.

The Plan seeks to make Türkiye a country that develops, produces, exports and reliably governs AI. Its stated principles are human-centredness, trustworthiness, ethical responsibility, digital sovereignty and sustainable development. It frames AI as both an economic and strategic capability, while recognising risks associated with algorithmic bias, privacy violations, cybersecurity threats, disinformation, intellectual-property disputes and labour-market change. It is a policy instrument rather than a comprehensive AI law, and its glossary states that concepts requiring legal binding force will be separately defined through legislation and secondary regulation.

Definitions

The Plan includes a non-binding glossary intended to explain policy and implementation terminology. It defines an Algorithmic Impact Assessment as a systematic process for analysing and documenting an AI system’s possible effects on individuals, society and institutions in advance. A Model Card is standard documentation summarising a model’s purpose, intended uses, training and data sources, performance metrics, known limitations, risks and recommended conditions of use. For high-impact systems, a public summary model card is to be used alongside a detailed technical file.

Key operational concepts include a high-impact AI system, meaning an application capable of directly affecting health, safety, fundamental rights or social welfare, including uses in health, education, employment, social assistance, credit assessment, biometric identification, law enforcement and justice. The glossary also defines a regulatory sandbox as a controlled, time-limited testing environment for innovative AI in regulated sectors. The proportional-risk approach means that obligations should increase with risk: low-risk applications receive lighter controls, while high-impact systems receive more extensive assessment.

Governance and Institutional Framework

The Plan establishes the National Artificial Intelligence Council as the highest strategic decision-making and coordination body for strategy, prioritisation, resource allocation, and risk and ethics principles. Chaired by the President and involving relevant ministries and affiliated institutions, the Council is to approve national priorities and annual implementation plans, resolve inter-institutional obstacles and competence disputes, direct major public and ecosystem investments, and publish an annual National Artificial Intelligence Policy Letter.

The Ministry of Industry and Technology’s Directorate General of National Technology and Artificial Intelligence serves as secretariat. A Programme Office is to prepare annual implementation plans, manage projects, operate the National AI Progress and Transparency Portal, and track action-level performance. The National AI Ethics Council will provide ethical guidance, practical principles and referral handling, but will not possess binding inspection or sanction powers; those remain with competent sectoral regulators. Technical security assessment capacity is to be strengthened under the coordination of the Cybersecurity Directorate in cooperation with TÜBİTAK Artificial Intelligence Institute.

Key Focus Areas

The Plan is organised around four axes, each containing four actions. “Fark Et” covers AI literacy, advanced talent, data ecosystems and a legal-ethical framework. “İstifade Et” covers a national computing portfolio, the “GPU for All” programme, public-sector AI transformation, and sectoral AI pilots and SME vouchers. “Üret” covers AI Growth Zones, research and growth funds, sectoral foundation models and exports, and physical AI and robotics. “Yönet” covers investment attraction, international AI diplomacy, regional capacity-building and safety evaluation with regulatory sandboxes.

Headline targets include training at least 1.5 million people in AI literacy in the first 12 months and five million citizens in total within 24 months; developing 10,000 advanced AI specialists and 100,000 AI application professionals; publishing at least 2,000 public datasets through the National Data Library; reaching at least 1 GW of AI and data-centre installed capacity by 2030; and mobilising at least USD 10 billion in predominantly private-sector AI and data-centre investment. The Plan also seeks more than TRY 1 trillion in economic value from combined public resources and private investment.

Implementation Framework

Implementation follows four principles: speed and clarity, market orientation, trust and responsibility, and phased implementation and prioritisation. The operational model is “scan-pilot-scale”: systematically identify use cases, run rapid pilots, and rapidly scale pilots that meet defined success thresholds. During 2026-2027, priority is given to setting up governance structures, piloting data spaces, initiating initial public-sector AI projects and clarifying investment processes for computing capacity. During 2028-2030, pilots are to be scaled, sectoral foundation models commercialised, international partnerships deepened and investment objectives delivered.

Public bodies are expected to receive a responsibility matrix identifying responsible, accountable, consulted and informed institutions for each action. Each ministry is to designate a senior official responsible for AI transformation and, following designation, prepare a three-year AI roadmap and annual targets within 12 weeks. The Plan targets allocating at least 2% of total public investment-programme budgets to AI projects. It also prioritises public procurement as a market-shaping tool, including digital badges for proven public-sector AI solutions and potential listing in the State Supply Office Tech Catalogue.

Monitoring and Evaluation

The National AI Progress and Transparency Portal is designed as the public single window for monitoring target, budget and timeline compliance. For each action, it is intended to display the responsible institution, sub-tasks, schedule, budget versus expenditure, indicators, status colour and critical dependencies. Personal data and trade secrets are excluded from public disclosure, while role-based internal management dashboards are contemplated for institutional use.

The Plan requires the first public progress report to be published on the Portal within 12 months of publication, followed by annual reports. Reports must identify implementation status, target deviations, obstacles and corrective measures, and each reporting cycle is to include an independent evaluation component with academic or international-organisation participation. Metrics will cover infrastructure, data, talent, adoption, trust and regulation, investment and exports, and international or regional impact. Annual updates may revise actions, targets and budget assumptions in response to technological developments and international regulatory trends.

Penalties, Liability, and Appeals

The Action Plan does not itself establish a general catalogue of administrative fines, civil damages, criminal penalties, appeal procedures or a standalone AI enforcement authority. It therefore does not impose a comprehensive private-sector penalty regime comparable to a dedicated AI statute. The Plan anticipates a future general legislative basis defining principles, competences and responsibilities, with implementation detail to be developed through secondary legislation by competent bodies.

Nevertheless, the Plan sets policy expectations for risk management. Low-risk applications are to be subject to simplified checklists; medium-risk applications are to require a short-form Algorithmic Impact Assessment; and high-impact systems are to require an expanded assessment, a public model-card summary and a detailed technical dossier supplied to supervisory authorities. High-impact systems should undergo standards-based technical safety and performance assessment before deployment and when major version updates occur. Binding oversight and sanctions remain matters for relevant sectoral regulators and applicable Turkish law, including personal-data, cybersecurity, consumer, health, financial and other sectoral regimes.

Relationship to Other Instruments

The Plan succeeds the National Artificial Intelligence Strategy (2021-2025), which entered into force through Presidential Circular No. 2021/18 published in the Official Gazette on 20 August 2021. It draws on experience from that strategy, the 2024-2025 Action Plan, the Turkish Grand National Assembly AI Research Commission Report and the National AI Policy Document of the Presidential Science, Technology and Innovation Policy Council. It is designed as an implementation-oriented roadmap rather than a replacement for sectoral legislation.

Data-related actions are to be coordinated with the Personal Data Protection Law No. 6698 (KVKK). The Plan calls for AI-oriented anonymisation standards, data-transfer conditions and secure-data-space conformity criteria to be developed in coordination with the Personal Data Protection Authority. It also refers to KVKK-GDPR alignment for automated decision-making explainability, personal-data processing in model training and cross-border data transfers. In intellectual property, the Plan proposes a Turkish Patent and Trademark Office AI Fast Track for AI technologies and AI-assisted inventions.

International Alignment

International alignment is a core component of the “Yönet” axis. Türkiye intends to contribute to AI governance and standard-setting through the OECD AI Principles, the Global AI Safety network, the United Nations, G20, ISO/IEC, ITU and UNESCO. The Plan seeks bilateral and regional technology alliances, data agreements and frameworks for sharing computing resources, subject to KVKK, international data-protection obligations, national security, trade-secret protection and reciprocity.

The Plan also envisages regional AI-capacity partnerships with the Organisation of Turkic States, the Organisation of Islamic Cooperation, D-8 countries, Gulf countries and neighbouring partners. It proposes common data centres, shared computing capacity, federated data access and a common Turkic-languages large language model spanning Oghuz, Kipchak and Karluk language families. By the end of 2027, the Plan targets at least five bilateral AI cooperation framework agreements, at least three memoranda with international institutions, and the first production version of the common Turkic-languages model.

Implementation Timeline

Compliance Checklist

Sources and References