The key facts

  • Kacır: 'We can delegate authority to AI, but not responsibility'
  • According to him, five companies control about 70% of the world's AI computing capacity
  • Turkey's 2026–2030 plan: training 5 million citizens, 10,000 AI experts, the BILGE model

Responsibility cannot be shifted to an algorithm

Turkey's Minister of Industry and Technology Mehmet Fatih Kacır stated that artificial intelligence must not be allowed to operate without clear accountability, and that technological progress must remain centered on human responsibility. He spoke on Saturday at the TRT World Forum panel 'Navigating Accountability in the Age of AI.' 'We can delegate authority to AI, but we cannot delegate responsibility,' the minister said, warning that increasingly autonomous AI systems raise serious questions about transparency, safety, and accountability.

Risks and the concentration of computing power

Among the risks Kacır named are algorithmic discrimination, cyber threats, deepfakes, and the concentration of computing resources in the hands of a small number of technology companies. He cited figures: more than 90% of notable advanced AI models in 2025 were developed by industry, while, by his estimate, just five companies control about 70% of the world's AI computing capacity. The minister called for human-centered, risk-based regulation: AI applications with significant consequences for people should receive stronger safeguards and genuine human oversight.

Turkey's plan for 2026–2030

Kacır also outlined Turkey's own goals. The National AI Action Plan for 2026–2030 is intended to place the country among the leaders in AI technology. It provides for teaching AI fundamentals to 5 million citizens, training 10,000 advanced specialists and 100,000 AI professionals, and opening at least 2,000 public datasets to researchers and entrepreneurs. By 2030, Turkey intends to expand data center capacity to at least one gigawatt and attract at least $10 billion in private investment in data centers, cloud computing, and AI infrastructure. The first version of the Turkish large language model BILGE, which, according to the minister, will strengthen technological independence, will soon become available to developers. Also announced were 'regulatory sandboxes' in priority sectors and an AI Ethics Board that will address privacy, discrimination, and child safety. In terms of international engagement, Kacır named the platforms through which Turkey is ready to participate in global AI governance: the G20, the OECD, UNESCO, the Organization of Turkic States, and the UN.