AI-Generated Content and Copyright Law: Legal Gaps and Future Perspectives

Generative artificial intelligence ("generative AI") has placed sustained pressure on the doctrinal foundations of copyright: authorship, originality, and enforceability. This Article compares the United States and the European Union, and identifies persistent regulatory gaps arising from (i) indeterminate thresholds for human creative control in AI-assisted outputs; (ii) liability misalignment across a multi-actor production chain (model developers, deployers, platform intermediaries, and end users); and (iii) evidentiary obstacles in proving copying or tracing protected expression through opaque training and inference processes. Using Thaler v.

Perlmutter, Andersen v. Stability AI, and the EU's text-and-data-mining ("TDM") opt-out architecture as illustrative reference points, the Article argues for a compliance-centered settlement: targeted disclosure duties, conditional safe harbors for commercial systems, and harmonized technical standards for dataset transparency, rights reservation, and output provenance.