Law 101 in the Age of AI. Human Authorship, Provenance, and Evidentiary Integrity

This paper examines how generative and assistive AI systems complicate long-standing legal assumptions about authorship, intention, evidentiary reliability, and accountability. It argues that where text is materially shaped by AI, courts and institutions may face increasing difficulty in determining whether that text can still be treated as a reliable proxy for human meaning, adoption, and responsibility. Focusing on evidence, procedure, legal attribution, and governance, the paper develops a framework for understanding how AI-mediated text may weaken confidence in authenticity, reliability, semantic continuity, and standards of proof. It further argues that existing safeguards such as metadata, audit logs, repositories, and model-governance frameworks, while important, do not by themselves fully resolve the upstream provenance problem where legal consequence depends on attributable human origin or accountable adoption. In response, the paper proposes the concept of a provenance substrate: a provenance-aware infrastructure designed to preserve clearer records of authorship, revision, semantic continuity, and the relationship between human input and machine intervention. The paper concludes that provenance assurance should be understood not merely as a technical feature or governance preference, but as increasingly important legal infrastructure for preserving evidentiary integrity, procedural fairness, accountable decision-making, and the human centre of law in an AI-mediated environment. Version note: This is Version 2.0, revised for academic review on 26 March 2026. It updates the original version first published on 7 December 2025.