Bogazici University: As artificial intelligence (AI) continues to evolve, its relationship with copyright law is fraught with uncertainty, sparking debates over fair compensation for creators and the limits of intellectual property rights.
According to Anadolu Agency, the 2nd Symposium on News Copyright and Artificial Intelligence in Media, hosted at Bogazici University and organized by Anadolu, highlighted the legal ambiguities surrounding AI-generated content. Guido Westkamp, an academic at Queen Mary University of London and an expert in intellectual property and comparative law, emphasized the unresolved legal issues in this domain.
AI systems are trained on extensive datasets, which often include copyrighted materials, raising concerns about fair compensation for creators. Yet, Westkamp noted that the legal framework remains murky. "It's a very topical question, and the answer is that at present, nobody knows," he stated.
Current copyright law includes exceptions allowing data mining, primarily intended for scientific research rather than AI training. "There is an exception in law for copyright that allows data mining. That wasn't made for AI," Westkamp explained, noting two exceptions: one for scientific research and another for general purposes, where authors can opt out by declaring their works should not be used for training.
A recent ruling by a lower court in Germany has added complexity to the discussion. "There is now a very recent decision, one of the first decisions on whether this data mining exception can apply to AI," Westkamp mentioned. The court ruled that the exception applies generally, even if data mining is used to train generative AI models, potentially eliminating remuneration concerns.
This ruling's future in higher courts remains uncertain. Westkamp suggested it could reach the Court of Justice for the EU, where balance-related questions will likely be addressed. He believes creators should have rights to remuneration and the option to opt out. "My own personal view is that there needs to be maybe a right to remuneration," he said.
The question of who owns AI-generated content remains unresolved. Westkamp argued that if the content is fully automated, no one owns it, as copyright law is based on human creativity. However, if AI is used merely as a tool, human authorship remains intact.
For media organizations using AI tools, adherence to copyright laws is an evolving challenge. Currently, transparency is key, with the EU's AI Act requiring clear records of works used in AI systems.
Westkamp sees AI as a fundamental test for intellectual property law in the digital age, raising philosophical questions about the relationship between humanity and robots. "The question of what does that have to do with fundamental rights?" he pondered, suggesting that these issues will shape legal debates in the coming years. "We will see in 20 years how things have turned out," he concluded.