Universiteit Leiden

nl en

Tech companies breach copyright when training AI tools

For some years, large tech companies have breached copyright law by using creative expressions from artists to train AI tools. Dirk Visser, Professor of Intellectual Property Law, spoke to ‘Financieele Dagblad’: 'Artists may lose the motivation to create new art.'

Tech companies such as Meta and Microsoft do not take copyright law very seriously. For two and a half years, photos, music, art, and books by artists, authors, and other creators have been used to train AI models. This has been done without permission or compensation for the original creators. Unfortunately, the legal frameworks in this area are outdated and insufficiently aligned with the issues related to AI and copyright.

This is especially painful for writers, musicians, illustrators and other artists who rely on making a living from their creations. If their work can easily be copied and without compensation, and it is not possible to distinguish robot art from real art, the question is whether people will still be willing to pay for the real thing, says Professor Visser. The next step, according to him, is that artists may lose the motivation to create new art.

More information?

Read the full atricle in FD (€, in Dutch)

This website uses cookies.  More information.