Web scraping: Jurisprudence and legal doctrines

Abstract

Web scraping is a technique that allows the extraction of online information and data to train Generative Artificial Intelligence (GenAI) systems. Although the use of deep learning algorithms to produce user-requested outputs (texts, images, music and code) based on models learned from vast data sets dates back a few decades, its use has become fundamental with the recent development of GenAI and has been accompanied by the emergence of the first legal disputes. Doctrine and jurisprudence are called upon to consider the legal consequences arising from the combination of web scraping and GenAI, often encountering inadequate and fragmented legislation. Laws and regulations vary significantly across different countries and regions, reflecting diverse priorities and legal approaches. However, while doctrine, regardless of the latitudes, agrees in condemning the illicit acts and abuses due not so much to the extraction method but to the use of the extracted data (where protected by intellectual property rights), jurisprudence (particularly in Europe and North America) has already had the opportunity to express divergent opinions in some leading cases.