The rapid growth of generative AI has intensified the debate over whether AI developers can legally use copyrighted material to train their models. According to the Council of Europe's European Audiovisual Observatory, AI systems are trained on vast collections of text, images, music, videos, and other creative works, many of which are protected by copyright. This raises fundamental questions about whether using these works for AI training requires permission from rights holders, whether existing copyright exceptions apply, and how creators should be compensated when their content contributes to AI development.
One of the key challenges is balancing innovation with creators' rights. In the European Union, the Text and Data Mining (TDM) exceptions allow certain uses of copyrighted material for AI training under specific conditions, but creators can opt out of commercial data mining. At the same time, the EU AI Act introduces transparency obligations for providers of general-purpose AI models, requiring them to implement copyright compliance policies and provide summaries of the data used for training. These measures aim to improve accountability while allowing AI innovation to continue.
The report also highlights the lack of global legal consistency. Different countries are adopting different approaches to AI training and copyright, leading to legal uncertainty for AI developers and content creators alike. Ongoing court cases around the world are examining whether AI training constitutes copyright infringement, whether it qualifies as fair use or other legal exceptions, and what level of transparency AI companies should provide about their training datasets. As AI becomes more widespread, these legal questions are expected to shape future copyright policies internationally.
The Council of Europe concludes that resolving these copyright challenges will require a careful balance between encouraging AI innovation and protecting intellectual property. Greater transparency, clearer licensing mechanisms, and harmonized legal frameworks could help reduce uncertainty while ensuring that creators are fairly recognized and compensated for the use of their works in AI training. As governments continue updating AI regulations, copyright is expected to remain one of the most significant legal issues affecting the future of generative AI.
