Turkey's Artificial Intelligence Action Plan (2026–2030): A Reading Through the Lens of an AI Startup

At the Turkey Artificial Intelligence Summit held on June 13, 2026, the Ministry of Industry and Technology shared the 2026–2030 Turkey Artificial Intelligence Action Plan with the public. The Plan sets out a comprehensive roadmap aimed at positioning Turkey not merely as a country that consumes technology in the field of artificial intelligence, but among those that develop, apply, and govern it.

As Mirleon AI, we wanted to assess what this Plan means for both our country and the domestic startups within the ecosystem, viewed through the lens of an AI startup developing AI-based decision support systems. After all, the opportunities this roadmap opens up are most relevant to young companies like ours—those that build technology from the ground up and aim for rapid growth.

The Backbone of the Plan: Four Pillars, Sixteen Actions

The Action Plan is built on four core pillars summarized in its motto: Discover, Benefit, Produce, Govern. Each pillar contains four priority actions, and the 16 actions in total form a holistic framework spanning from awareness to global leadership.

• Discover: Building awareness, talent, the data ecosystem, and the rules.
• Benefit: Establishing and using the infrastructure; adopting it across the public sector and industry.
• Produce: Producing, financing, developing models, and ensuring their security.
• Govern: Attracting global investment, conducting international diplomacy, and becoming a regional leader.

The Plan treats the artificial intelligence ecosystem as mutually reinforcing layers: energy, computing hardware, infrastructure, models/data, and applications form the core backbone; while talent, financing, and the regulatory framework constitute the cross-cutting components that feed all the layers.

Key Targets

The Plan's concrete and measurable targets clearly indicate which areas will receive priority:

• Human capital: Training 10,000 advanced AI specialists and 100,000 AI application professionals; basic training for 5 million citizens within two years; literacy workshops across all 81 provinces.
• Computing power: At least 1 GW of installed data center capacity by 2030 and a national computing backbone powered by low-carbon energy.
• Investment: Mobilizing at least USD 10 billion in investment, predominantly from the private sector; a projected economic value exceeding 1 trillion TL across the Plan as a whole.
• Data: Opening access to at least 2,000 public datasets in machine-readable, standardized, and openly licensed form.
• Access: Providing computing capacity to startups, researchers, and SMEs under fair conditions through the "GPU for Everyone" program; allocating a significant portion of the credits to be distributed to tech ventures.

What Does It Mean for AI Startups?

A closer look at the Plan reveals numerous mechanisms that directly concern early-stage AI startups. For a venture where access to capital, computing power, and data is the most critical constraint, these items are decisive:

Access to computing power. The cost of computing, one of the biggest obstacles facing an AI startup, is addressed through the "GPU for Everyone" program (National AI Research Resource). The program aims to provide GPU-hour credits through accredited data centers, TRUBA, and the national computing portfolio, to allocate at least 2 million GPU-hours annually in its first phase, and to reserve at least 40% of these credits for tech ventures. For startups competing with limited resources, like us, this is a game-changing lever.

Access to data. Access to high-quality, anonymized, and licensed datasets through the National Data Library and Data Spaces is critical for any startup developing data-driven solutions. Roadmaps are being planned for priority sectors such as healthcare, agriculture, defense, and e-commerce.

A support pipeline from idea to product. The Sectoral AI Pilot Programs are designed as a multi-stage competition-and-support mechanism that carries applicant teams from the idea stage all the way to a field-ready product, offering increasing support at each stage (idea, prototype, field trial, scaling).

The investment and financing ladder. The Plan establishes a two-tiered fund structure spanning from basic research to startup scaling: the National AI Research Fund, which supports early-stage projects, and an advanced-stage growth fund that accelerates the Series A/B rounds of maturing ventures. Opening a dedicated investment window for AI startups within the advanced-stage fund is a direct commitment to startups. When combined with the USD 10 billion private-sector investment target, a substantial flow of capital for startups is anticipated within the ecosystem.

Access to the public market. Allocating at least a 2% share to AI projects from the public investment program budget, and having the public sector serve as the first customer for domestic solutions, offers one of the most valuable things for a new venture: a proven reference and predictable demand. Moreover, the "digital badge" and export presentation packages to be awarded to solutions successfully deployed in the public sector pave the way for domestic startups to enter the global market backed by a strong reference.

Talent pool and growth zones. The target of training 10,000 advanced specialists and 100,000 application professionals will nourish the skilled human capital that startups need most. AI Growth Zones and Regional Centers of Excellence, meanwhile, provide ready-made infrastructure for rapid prototyping, testing, and productization.

Sectoral Models and a Focus on Turkish

The Plan emphasizes that global competition is not limited to large language models alone; it draws attention to the importance of sector-specific foundation models that can be developed using countries' domain-specific data advantages in areas such as healthcare, drug discovery, genomic research, disaster management, energy, agriculture, and new materials.

The AI-based decision support systems and healthcare-oriented solutions that Mirleon AI is working on fit precisely within this framework. The wealth of high-quality data produced by Turkey's healthcare infrastructure offers a strong foundation for a competitive position in sector-specific models. In parallel, the continuation of efforts on a Turkish large language model to strengthen digital sovereignty, and the development of a joint Turkic Languages Large Language Model covering the Oghuz, Kipchak, and Karluk language families within the framework of the Organization of Turkic States, are also among the Plan's notable highlights.

A Trustworthy and Ethical Framework

In our view, one of the most important aspects of the Plan is its adoption of a regulatory approach that does not slow innovation but does make safe scaling possible. A national legal-ethical framework inspired by the European Union's proportional risk approach foresees graduated obligations according to the level of risk.

In addition, regulatory sandboxes are being established so that innovative solutions can be safely tested in regulated sectors such as finance, healthcare, energy, mobility, and telecommunications; the transition of successful pilots into permanent legislation or public procurement is being facilitated. This approach is a significant gain in terms of both predictability and reliability for teams working in high-sensitivity domains.

Conclusion: Being on the Producing Side

The 2026–2030 Turkey Artificial Intelligence Action Plan offers a holistic approach that touches every layer of the AI ecosystem, from human capital to computing infrastructure, and from data to regulation. The core proposition driving the Plan is clear: to move Turkey among the countries that produce and govern artificial intelligence, rather than merely use it.

As Mirleon AI, we believe the opportunities this roadmap opens up hold concrete prospects for domestic technology-producing startups like ours. As an AI startup, with an understanding that sees artificial intelligence as a tool that protects human dignity, strengthens decisions, and creates value, we will continue to stand on the producing side of this transformation.