What happened?

Turkey's 2027-2029 Medium Term Programme has been published. Alongside economic and financial headings, the 81-page document sets targets for technology, energy and the defence industry. Artificial intelligence, semiconductors, robotics, batteries and autonomous systems are among the areas singled out.

What stands out is that the programme treats these areas together rather than separately: it aims to strengthen research, design, manufacturing and commercialisation so they reinforce one another. Increasing strategic cooperation between research institutions, universities and the private sector is another heading.

What is there on the AI side?

  • Developing high-performance computing and supercomputer infrastructure, and expanding the capacity of the TRUBA supercomputer.
  • Establishing the standards needed for public institutions' data to be used in AI applications — target date the second quarter of 2027.
  • Developing large language models with high proficiency specifically in Turkish.
  • Building domestic generative AI solutions on those models, customisable to public institutions' needs.

Chips and autonomous systems

Semiconductors are listed as a strategic area. The programme targets modern domestic facilities capable of producing the semiconductor chips used in industrial manufacturing, and plans for the chips used in identity cards and passports to be produced in Turkey.

Autonomous vehicles and systems sit directly among the priority R&D areas. Next-generation transport technologies, advanced materials, robotics and energy storage that would support them are also on the strategic list. Spreading AI-supported and autonomous production systems through industry is likewise planned.

How broad is the "deep technology" list?

Under deep and advanced technology, the programme lists: artificial intelligence, cybersecurity, autonomous systems and vehicles, quantum technologies, hypersonic systems, next-generation materials, nanotechnology, biotechnology, sensor technologies, chips, batteries, nuclear, renewable energy, blockchain and robotics.

The breadth raises a question: declaring priority across fifteen areas at once can weaken the notion of priority itself. The programme mostly does not give a ranking among them, an allocated budget line, or a measurable target — how many facilities, how much capacity, by what date. The exception is the second quarter of 2027 given for public data standards.

Energy and defence

On energy, investments raising solar and wind capacity get priority, and spreading battery storage projects is targeted so variable generation can be integrated into the system. Increasing nuclear capacity and raising the domestic content of plant equipment are also in the programme. A national plan will be prepared for high-capacity charging stations to support the electrification of heavy commercial vehicles.

The most striking item under defence is transfer: the programme aims to open the way for technology, experience and know-how developed for defence to move into civilian fields such as transport, health, AI and security. Extending the demand-creation and long-term ordering approach used successfully in defence to health technologies, rail systems and electronics is also planned.

What to watch

A programme document is a statement of intent; what can be measured is its calendar and its budget. In this one, the only concrete AI item with a date is public data standards. The Turkish large language model target has neither a date nor a capacity measure — nor does it say who would build the model, on what compute, or trained on which data.

A similar gap shows in the space technologies section: there are no clear targets for satellites, launch systems or space missions. What to watch in the coming period is not the list itself but how many of its items turn into a plan with a date and a budget.