Παραγωγική τεχνητή νοημοσύνη / γλωσσικά μοντέλα (ChatGPT κ.α.) και προστασία προσωπικών δεδομένων

Ράσα, Άντζελα

Generative AI / language models (ChatGPT etc.) and data protection

Master Thesis

Author

Ράσα, Άντζελα

Date

2024-01

Abstract

This paper aims to examine the peculiar relationship between modern artificial intelligence technologies (AI), such as generative AI and large language models, and privacy protection. While generative AI has introduced significant innovations and conveniences in human communication and business practices, its extensive use of data, including personal data, raises important legal and ethical issues concerning compliance with existing data protection laws and societal ethical standards. The thesis starts with a historical overview of the development of AI, an attempt to define the concept of AI and an analysis of AI techniques and approaches. A necessity for understanding productive AI is not only the prior understanding of traditional AI but also the understanding of Big Data. Next, productive AI is examined through the structure and mode of operation of big language models such as ChatGPT. In the second chapter of this paper, the legal framework for data protection, specifically the General Data Protection Regulation (GDPR) and its application to language models is discussed. Its conflicts with respect to the practical operation of large language models are listed and relevant ways for AI to comply with the current legislative framework are suggested. Then, the legislative framework for AI is mentioned, with emphasis on newly published AI ACT and its relation to the GDPR. Finally, the ethical issues of the use of AI are mentioned, as well as some landmark decisions of the Court of Justice of the European Union (CJEU), which, in the absence of case law on productive AI, which can be applied analogously to generative AI.

Postgraduate Studies Programme

Δίκαιο και Τεχνολογίες Πληροφορικής και Επικοινωνιών (MSc in Law and Information and Communication Technologies)

Department

Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών Συστημάτων

Number of pages

Language

Greek

URI

https://dione.lib.unipi.gr/xmlui/handle/unipi/17429

Collections

Τμήμα Ψηφιακών Συστημάτων

Show full item record

Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα