The Fight for Clearer Laws on Human Authorship
When artificial intelligence participates in making a work, what did the human actually author?
Georgetown University Library has been helping document that question through its Artificial Intelligence (Generative) Resources research guide, including its discussion of AI, authorship, and copyright.
Georgetown University Library — Artificial Intelligence (Generative) Resources
Last Updated: September 17, 2026, 3:57 PM
Their work is a useful starting point for creators trying to understand a legal landscape changing almost as quickly as the technology itself.
AI use does not automatically erase copyright.
The U.S. Copyright Office has drawn an important distinction between using AI as a tool and allowing AI to become the source of the expressive work itself.
Human-created expression can still receive copyright protection when AI is involved. Human selection, arrangement, modification, editing, and other creative contributions may also be protected.
What generally cannot be protected is purely machine-generated expression where there is insufficient human control over the final result.
The legal question is not simply: Did you use AI?
It is: Where did the human creative decisions occur?
A prompt is not necessarily authorship.
A filmmaker might type: Woman standing beneath a streetlight. Rain. 50mm lens. Slow push in. Moody lighting.
That communicates creative intent. But the AI may still determine the woman's appearance, composition, background, lighting details, movement, and dozens of other elements.
The human directs. The system interprets.
That gap between intention and execution is increasingly important to copyright law. In Thaler v. Perlmutter, an AI-generated image was submitted for copyright registration with the machine itself listed as the author. The courts rejected that claim, and the Supreme Court declined to hear the case in March 2026.
The principle remains: under current U.S. law, authorship begins with a human being.
Filmmaking makes the distinction easier to see.
Directors have never personally created every element of a film. Actors perform. Cinematographers operate cameras. Editors cut footage. Visual-effects artists build images. Software stabilizes, tracks, composites, and increasingly automates parts of production.
Yet authorship remains because humans make creative decisions: use this performance, reject that shot, change the timing, return to version three, keep this character consistent, approve this frame.
AI does not eliminate those decisions. It may make documenting them more important.
A production history can preserve something a final generated file cannot:
Intent → Generation → Judgment → Revision → Approval
That record does not automatically create copyright protection, but it can help show where human creative control actually existed.
Preserve the human decision trail.
AI makes the video. Atlas Observe remembers the production.
This is one of the ideas behind Atlas Observe: preserving the human decision trail behind AI-assisted production rather than treating generation alone as authorship.
This matters to small businesses too.
The issue is not limited to filmmakers or large studios. Small businesses are already using generative AI for advertising, social content, product imagery, voiceovers, and commercial video.
In my work at Ironside, AI can expand what is possible within a small-business production budget. That is valuable. But it creates a practical question for agencies and clients:
What exactly is the business paying to own?
If someone enters a prompt and hands over the first generated result, the human-authored contribution may be relatively limited. If a creative team develops the concept, writes the campaign, directs the visual language, selects references, rejects generations, edits the results, and assembles the finished commercial, the human contribution becomes much easier to identify.
Businesses may need to start asking not only whether AI was used, but who made the creative decisions and how the finished work was developed.
The fight now is for clarity.
Creators still need clearer standards. Where does assistance end and authorship begin? How much expressive control is enough? How should agencies, filmmakers, and businesses document their contributions?
These are no longer theoretical questions.
Georgetown University Library's continually updated AI research guide is helping frame that conversation, while the Copyright Office and courts continue defining the boundaries.
As AI generation becomes cheaper and more abundant, authorship may become more valuable rather than less.
Eventually the question will stop being: Can AI make this?
It will become: Who authored it?
Creators deserve clearer rules for answering that question.
Reference: Georgetown University Library, Artificial Intelligence (Generative) Resources. Last Updated September 17, 2026, 3:57 PM.
This article discusses developing U.S. copyright law and is commentary, not legal advice.