Book

Artificial intelligence has outpaced the legal categories built to contain it. This collection confronts two of the most urgent fault lines in that unraveling: the theft of human likeness through deepfakes and voice-cloning, and the wholesale ingestion of human writing to train the machines that increasingly speak in our place. The first essay argues that existing law - fractured across privacy, defamation, and intellectual property - was never built for a world where identity itself can be synthetically reproduced. It proposes a new legal category: synthetic identity rights, grounded in autonomy and dignity, and tested against regulatory approaches in the European Union, the United States, and Nigeria. The second essay turns to Bartz v Anthropic PBC, the landmark 2025 ruling that found AI training to be fair use while treating piracy as an unforgivable shortcut. Read against Nigeria's Copyright Act 2022, the case reveals a striking asymmetry - one that leaves African rightsholders more protected, and African AI developers more exposed, than their American counterparts. Together, these essays make a single argument: that the law has been asking the wrong question. Not whether synthetic content is true or false, transformative or not - but whether a person's identity, or a person's words, have been taken without their say. For scholars, practitioners, and policymakers navigating the collision between artificial intelligence and the rights it displaces, this book offers both a diagnosis and a roadmap.