Generative artificial intelligence (AI) poses significant challenges to intellectual property (IP) rights and the knowledge economy. The article explores two key policy questions: how to compensate data creators for AI model training and who owns AI-generated outputs. These issues are central to IP law, which traditionally rewards human creativity. While some jurisdictions, like the UK and Singapore, have introduced exceptions for text and data mining, the broader implications of these choices are complex. The article draws parallels with the music industry's past struggles with piracy, suggesting that similar litigation and legislation may help navigate current uncertainties. It also highlights the growing market for "legitimate" AI models that respect copyright and provenance. The article discusses the legal status of AI-generated content, noting that in most jurisdictions, automatically generated text does not receive copyright protection. However, content that is edited or curated may be owned by the person doing so. The UK's approach to "computer-generated" works, which recognizes the person who arranged the creation as the author, is one possible solution. The article also notes that AI-generated content may not be eligible for copyright protection, but edited content may still be owned by the editor. The article argues that while AI has the potential to transform the arts and knowledge economy, strict IP protection could hinder innovation, while insufficient protection could undermine the viability of the arts sector. It suggests a middle ground, such as reduced protection terms and other limitations, to balance these concerns. The article also discusses the need for transparency in AI development and deployment, including disclosing data origins and the relative contribution of AI to new works. The article concludes that while AI has the potential to transform the arts and knowledge economy, the legal and economic implications are complex. It emphasizes the need for balanced policies that protect human creativity while encouraging innovation. The article also highlights the importance of recognizing the intellectual input that goes into data training AI models and the need for transparency in AI development and deployment.Generative artificial intelligence (AI) poses significant challenges to intellectual property (IP) rights and the knowledge economy. The article explores two key policy questions: how to compensate data creators for AI model training and who owns AI-generated outputs. These issues are central to IP law, which traditionally rewards human creativity. While some jurisdictions, like the UK and Singapore, have introduced exceptions for text and data mining, the broader implications of these choices are complex. The article draws parallels with the music industry's past struggles with piracy, suggesting that similar litigation and legislation may help navigate current uncertainties. It also highlights the growing market for "legitimate" AI models that respect copyright and provenance. The article discusses the legal status of AI-generated content, noting that in most jurisdictions, automatically generated text does not receive copyright protection. However, content that is edited or curated may be owned by the person doing so. The UK's approach to "computer-generated" works, which recognizes the person who arranged the creation as the author, is one possible solution. The article also notes that AI-generated content may not be eligible for copyright protection, but edited content may still be owned by the editor. The article argues that while AI has the potential to transform the arts and knowledge economy, strict IP protection could hinder innovation, while insufficient protection could undermine the viability of the arts sector. It suggests a middle ground, such as reduced protection terms and other limitations, to balance these concerns. The article also discusses the need for transparency in AI development and deployment, including disclosing data origins and the relative contribution of AI to new works. The article concludes that while AI has the potential to transform the arts and knowledge economy, the legal and economic implications are complex. It emphasizes the need for balanced policies that protect human creativity while encouraging innovation. The article also highlights the importance of recognizing the intellectual input that goes into data training AI models and the need for transparency in AI development and deployment.