This chapter examines the most pertinent issues facing copyright law as it encounters increasingly sophisticated artificial intelligence (AI). It begins with a few introductory examples to illuminate the potential interactions of AI and copyright law. Section 1 tackles the question of whether AI-generated works are copyrightable in Canada and who, if anyone, might own that copyright. This involves a doctrinal discussion of “originality” (the threshold for copyrightability) as well as reflections on the meaning of “authorship,” and concludes with the suggestion that autonomously generated AI outputs presently (and rightly) belong in the public domain. Section 2 turns to consider issues of copyright infringement. First, it addresses the law in respect of AI inputs (the texts and data used to train AI systems, which may themselves be copyrightable works) and highlights the need for greater limits and exceptions to ensure that copyright law does not obstruct best practices in the development and implementation of AI technologies. It then examines the matter of potentially infringing AI outputs (which may, of course, resemble copyright-protected, human-created works), identifying current uncertainties around independent creation, agency, and the allocation of liability. Section 3 addresses the deployment of AI in automated copyright-enforcement, emphasizing its increasingly critical role in shaping our online environment and citizens’ everyday encounters with copyright enclosures. The chapter concludes with reflections on the risks and opportunities presented by AI in the copyright context, and identifies key gaps and questions that remain to be answered as copyright law and policy adjust to evolving AI technologies.
Craig, Carys, "AI and Copyright" (2021). All Papers. 358.