Algoritmanın Ötesinde: Avrupa Birliği Hukukunda Yapay Zekânın Fiziksel Temellerinin Yönetimi
International Congress on Artificial Intelligence and Interdisciplinary Applications, İstanbul, Türkiye, 21 - 22 Ekim 2026, (Yayınlanmadı)
- Yayın Türü: Bildiri / Yayınlanmadı
- Basıldığı Şehir: İstanbul
- Basıldığı Ülke: Türkiye
- Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
- Anadolu Üniversitesi Adresli: Evet
Özet
ABSTRACT
When lawyers talk about artificial intelligence,
they usually talk about algorithms, data, and risk. Much less is said about the
machines and physical infrastructure that make artificial intelligence
possible. AI ultimately depends on chips and computing infrastructure; it is
digitally constituted but materially grounded. This paper argues that this
physical side should be understood as part of AI governance. To make this case,
the paper reads the EU Artificial Intelligence Act together with three
instruments rarely discussed alongside it: the Chips Act, the Cyber Resilience
Act, and the NIS2 Directive. In fact, the AI Act itself already looks beyond
the algorithm. It defines AI systems in terms of outputs that can influence
physical as well as virtual environments. It also uses the amount of
computation employed in training a general-purpose AI model as a criterion for
identifying models with systemic risk. The Chips Act, for its part, addresses
the semiconductor capacity on which such computing depends. The Cyber
Resilience Act and NIS2 help secure the products and networks on which AI
systems rely. Read together, these laws show that the EU has already begun to
regulate AI as both software and infrastructure, though piece by piece and
across separate instruments. Recognizing this connection helps explain where EU
technology law is heading: towards a closer link between regulating AI,
securing what it runs on, and building Europe’s industrial capacity.
Keywords: Artificial intelligence; EU AI Act; EU Chips
Act; semiconductors; physical infrastructure; technology governance.