AI regulation in Turkey (2026)

A new bill submitted to the Grand National Assembly on 8 January 2026 proposes adding a misdemeanor to KVKK Law 6698 that would impose fines of up to 5% of the prior year's global turnover on social media and digital platforms that allow sharing of AI-generated audio, written, or visual content without the data subject's consent. It represents Turkey's first move toward platform-level financial liability specifically tied to AI-output harms.

Turkey's Personal Data Protection Authority published a 63-page 'Guideline on Generative AI and the Protection of Personal Data (in 15 Questions)' on its official website, covering training-data legality, explicit consent, cross-border transfers, transparency obligations, and risks such as deepfakes and hallucinations. It is considered one of the most detailed official generative-AI guidance documents issued by any data protection authority worldwide.

Bill Esas No. 2/3358, filed in the Grand National Assembly on 7 November 2025, proposes targeted amendments across five existing statutes — Internet Law 5651, Penal Code 5237, KVKK 6698, Electronic Communications Law 5809, and Cybersecurity Law 7545 — to embed AI governance obligations: statutory AI-system definitions, mandatory deepfake labelling (TRY 500,000–5,000,000 fines), six-hour content takedown windows, expanded BTK emergency blocking powers, and criminal liability for developers whose systems are weaponised. Turkey chose an amendment-based approach rather than a standalone AI Act.

Ankara's Chief Public Prosecutor opened a criminal investigation into xAI's Grok chatbot after it generated content insulting President Erdoğan and Mustafa Kemal Atatürk; a Turkish court ordered BTK to block the offending outputs. Turkey became the first country to impose regulatory measures specifically on an AI chatbot's generated content, demonstrating that AI-generated speech is directly actionable under Law No. 5651 without any dedicated AI statute.

The Personal Data Protection Authority published an explanatory note specifically addressing chatbot systems, clarifying that KVKK obligations fully apply when any chatbot processes personal data, requiring Data Protection Impact Assessments, user notice describing data retention and third-party sharing, and privacy-by-design integration. It was the Authority's first guidance document focused exclusively on the conversational-AI use case.

The Presidency Digital Transformation Office and Ministry of Industry and Technology released a mid-term action plan updating the National AI Strategy with 71 concrete measures across six strategic pillars, including development of Turkish-language large language models, expanded public AI infrastructure, cross-ministerial coordination, and enacting regulations to accelerate socioeconomic integration. The plan anchors Turkey's AI ambitions within the 12th National Development Plan.

An eight-article draft AI law was submitted to the TBMM in June 2024, proposing foundational definitions, risk-based obligations, mandatory AI-content labelling for deepfakes, personality-rights protections with six-hour takedown duties, and designating BTK and KVKK as co-enforcement authorities. Although concise, it was Turkey's first attempt at a dedicated AI statute and launched the parliamentary debate that shaped the more detailed November 2025 bill.

The Presidency Digital Transformation Office published voluntary ethical guidelines for AI development and deployment, centred on human-centredness, transparency, accountability, non-discrimination, and privacy, aligning with OECD AI Principles and Council of Europe frameworks. The guidelines were non-binding but represented Turkey's first official statement of AI ethics norms and explicitly flagged the need for future binding regulation.

Turkey's Personal Data Protection Authority published its first dedicated AI guidance, setting sector-spanning expectations on privacy-by-design, data minimisation, accountability matrices throughout the AI project lifecycle, and the need for impact assessments for high-risk processing. Published directly on kvkk.gov.tr, this document became the operative data-protection standard for AI developers in Turkey in the absence of any AI-specific statute.

Presidential Circular No. 2021/18, published in Official Gazette No. 31574, established Turkey's first National Artificial Intelligence Strategy, setting 2025 targets of raising AI's share of GDP to 5%, training 50,000 AI specialists, and organising 24 objectives and 119 measures under six strategic priorities including a dedicated regulatory and ethics pillar aligned with EU frameworks. It remains the foundational policy document underpinning all subsequent AI governance actions in Turkey.

Law No. 6698, published in Official Gazette No. 29677 on 7 April 2016, established Turkey's comprehensive data protection framework modelled on EU Directive 95/46/EC, created the independent Personal Data Protection Authority (KVKK), and introduced rights over automated decision-making affecting individuals. As the primary law governing personal-data processing in AI systems — including training datasets — it serves as the legal bedrock for all AI-related data governance obligations in Turkey.