Copyright Status and Ownership of AI-Generated Text

Determine whether text produced by large language models is protected by copyright and identify the rightful owner of such copyright, if any.

Background

Beyond the legality of training data, the authors note uncertainty regarding the copyright status of outputs generated by LLMs. Some platform terms assign ownership to users, but this may not resolve the underlying legal status of AI-generated text.

Establishing clear legal guidance on whether AI-generated outputs are copyrightable—and who owns them—has practical consequences for scientific writing, licensing, attribution, and reuse.

References

It is also unclear if text generated by LLMs is protected by copyright and who owns the copyright. In some cases (e.g., OpenAI), the service agreement grants the copyright to the user, but the issue is still open from a legal standpoint.

— What is the Role of Large Language Models in the Evolution of Astronomy Research?  (2409.20252 - Fouesneau et al., 2024) in Section: Ethical and Legal Concerns — General Concerns

Whether a reader may legally reuse the code is, for many of these systems, unclear.

— Science Done on a Machine by a Machine: AI Agents in Computational Chemistry  (2608.18508 - Dral et al., 19 Aug 2026) in Section "Where do we stand now and what's the end game?"