The comparison between generative artificial intelligence and Napster is useful only when it is used to examine the construction of a new distribution system rather than to predict an identical legal outcome. Napster made the circulation of digital music socially ordinary before institutions had produced a credible agreement among listeners, artists, record companies and network operators. Closing the platform did not eliminate either the technical capacity or the public appetite it had exposed; streaming later converted access into a licensed interface, but also created new concentrations of power and remuneration structures that musicians continue to contest. Generative AI has likewise normalised a technical possibility before settling the conditions under which value should return to the cultural materials that made it possible. The parallel is not exact: training a model is not identical to sharing a music file, generated outputs vary in their proximity to sources, and legal systems already contain exceptions for analysis, research or text-and-data mining. Yet both moments reveal how quickly an informal practice can become an infrastructural expectation. Users begin to experience immediate access as natural, investors finance the intermediaries capable of providing it, and by the time rights are negotiated, the interface through which culture will be encountered may already have been fixed. The WIPO discussion of an AI “Napster moment” is valuable because it shifts attention from a single infringement dispute to the wider reorganisation of music production, metadata, licensing and royalties. FormatAdmission determines which claims can enter that negotiation. Large companies arrive with logs, lawyers, technical standards and measurable market positions, while dispersed authors must often prove not only ownership but the presence and effect of their work inside systems they cannot inspect. A fair settlement will therefore require more than individual lawsuits or small opt-out buttons. It may include collective licensing, standard dataset declarations, auditable rights registries, differentiated rules for commercial and scientific models, revenue participation where direct substitution is demonstrable, and public-interest access for archives and research. The Napster lesson is not that piracy inevitably becomes subscription, but that legal conflict tends to stabilise around a new commercial container. That container may solve the inconvenience experienced by consumers while leaving the distribution of value structurally unequal. The present opportunity is to avoid treating the first workable licensing model as the final cultural settlement. The decisive question is not whether artificial intelligence will have a Napster moment, but whether the arrangements that follow will merely legalise concentration or construct a more plural economy of access, attribution and cultural production.
World Intellectual Property Organization (2025). Artificial Intelligence: Is This the Industry’s Next Napster Moment? WIPO Magazine Music Special.
U.S. Copyright Office (2025). Copyright and Artificial Intelligence, Part 3: Generative AI Training.
Stober, S. and Dornis, T. W. (2026). Generative AI Training and Copyright Law.
Shanklin, G. et al. (2025). The Case for Contextual Copyleft.
