Generative AI systems are trained using vast amounts of data, often taken from sources in the public domain that may be protected by copyright or other intellectual property rights. So could training a generative AI system using publicly accessible copyright work constitute an infringement? And could the output infringe?
Our global guide has just been