Provenance, Authenticity, and Copyright Enforcement in the Age of Synthetic Media

Abstract

Synthetic media is no longer a theoretical concern. We are living through a quiet but seismic shift in how creative work is made, and the law is struggling to keep pace. Deepfakes, AI-generated images, fabricated audio, algorithmically written text. It is everywhere, and the intellectual property frameworks built over decades were simply not designed with any of this in mind.This paper takes a hard look at that mismatch. Copyright law has always assumed a human being at the centre of the creative act, someone who makes choices, brings perspective, and leaves something of themselves in the work. Generative AI scrambles that assumption entirely. Drawing on case law from the United States, the United Kingdom, and the European Union, alongside the EU AI Act 2024, the paper examines what happens to authorship, originality, and infringement when the "author" is a neural network trained on millions of works it may never have had permission to use.We also look beyond the courtroom at the technical tools proposed as practical fixes, content provenance standards, digital watermarking, blockchain-based attribution and ask honestly whether they are equal to the task.The conclusion is that tinkering around the edges will not be enough. What is needed is a fundamental rethink of authorship theory, liability allocation, and cross-jurisdictional cooperation. The paper closes with concrete policy recommendations aimed at protecting original creators without closing the door on genuine technological possibility.