Türkiye targets $10 bln in AI infrastructure investment

ISTANBUL

Türkiye aims to attract at least $10 billion in investment led by the private sector into data centers, cloud computing and artificial intelligence infrastructure by 2030, Industry and Technology Minister Mehmet Fatih Kacır said on Nov. 7.

Speaking at the TRT World Forum in Istanbul, Kacır outlined targets under the country’s 2026–2030 Artificial Intelligence Action Plan, including raising installed data center capacity to at least 1 gigawatt.

The plan also calls for AI literacy training for 5 million people, alongside the training of 10,000 advanced AI specialists and 100,000 professionals able to apply the technology in sectors including manufacturing, health care, agriculture and finance.

Kacır said the first version of BİLGE, a domestically developed Turkish-language large language model, would soon be made available to developers, researchers and startups.

The model is intended to support generative AI applications suited to Turkish and local needs, allowing companies to develop products with less reliance on foreign technology, he said.

The government also plans to provide secure access to at least 2,000 public-sector datasets to help researchers and entrepreneurs develop applications addressing practical problems.

Kacır said Türkiye was supporting AI and data center investment through its HIT-30 High Technology Investment Program.

Around 9,000 researchers have used the national supercomputer infrastructure, while Turkish researchers and startups have received free access to European supercomputing resources worth $100 million through the Digital Europe Program and the EuroHPC Joint Undertaking.

Alongside the investment and training targets, Kacır outlined plans for oversight of AI applications, saying greater autonomy for machines must not weaken human accountability.

“We can delegate authority to AI, but we can never delegate responsibility,” he said.

Regulation should reflect the risks associated with each application, Kacır said.

A system used to improve a factory production line should not face the same requirements as software making medical diagnoses or decisions affecting access to public services.

High-risk applications would require stronger safeguards, assessments before deployment and meaningful human oversight, he said.

The government also plans to establish regulatory sandboxes in priority sectors, allowing entrepreneurs and researchers to test new technologies under defined rules.

A planned AI ethics board would guide developers and institutions on personal data privacy, algorithmic fairness and children’s digital safety.