The US Supreme Court Just Ruled AI-Generated Art Cannot Be Copyrighted. What This Means for Every Developer and Creator.

The US Supreme Court has declined to review the appeal on AI-generated art copyright — making the lower court ruling final and binding: AI-generated artwork is not eligible for copyright protection in the United States.

This is not a technicality. It is a fundamental ruling that reshapes how developers, designers, and businesses can use, sell, and protect AI-generated content.

What the ruling actually says

The legal chain: the Copyright Office rejected copyright registration for AI-generated images, ruling that copyright requires human authorship. That was challenged by Stephen Thaler, who sought copyright for an image generated entirely by his AI system DABUS with no human creative input.

Federal courts upheld the Copyright Office: human authorship is a requirement for copyright. AI systems cannot be authors. Content generated without meaningful human creative input does not qualify.

The Supreme Court declining to hear the appeal makes this settled US law. No further appeals process exists.

What counts as human authorship — and what does not

The ruling does not make all AI-assisted art uncopyrightable. The key is creative input and control.

Content that likely retains copyright protection:

• An artist who uses AI as a tool while making significant creative decisions — choosing outputs, editing, combining elements, directing through extensive prompt engineering
• AI-generated images substantially modified or curated by a human into a larger creative work

Content that likely has no copyright protection:

• A single image generated from a simple text prompt with minimal human input
• Fully automated content pipelines with no human selection
• AI-generated music, text, or video where the human only provided the initial request

The line is not precisely defined. Legal ambiguity will persist for specific cases. But the principle is clear: less human creative input equals a weaker copyright claim.

What this means for developers

If your product generates AI content and delivers it to users:

1. Your users cannot copyright fully AI-generated outputs. If they plan to sell or license that content with copyright protection, they may be making a legally incorrect assumption — update your documentation.

1. Competitors can legally copy AI-generated content you publish without substantial human modification. Those assets are effectively in the public domain.

1. Training data benefit: using AI-generated images to fine-tune your own model is legally convenient — no rights to clear on public domain content.

What this means for designers

A designer who uses Midjourney to generate a base image and substantially modifies it in Photoshop has a credible copyright claim on the final work. A designer who generates an image and delivers it to a client with minimal modification is delivering something with no copyright protection.

The commodity is the output. The value is the curation, modification, and brand context surrounding it — and those remain protectable.

The international dimension

The US ruling does not bind other jurisdictions:

• The EU Copyright Directive allows protection where a human made "creative choices" — interpretation varies by member state
• The UK has a specific provision for computer-generated works (Section 9(3) CDPA)
• China granted copyright to AI-generated images in a 2023 Beijing court ruling

Developers building globally need jurisdiction-specific legal advice, not just the US baseline.

Practical steps

• Review any product that delivers AI content to users and clarify copyright status in documentation
• If your value proposition includes copyright ownership, document human creative input carefully
• Do not rely on copyright to protect AI-generated marketing assets from being reproduced
• If you encourage human modification workflows in your product, that is now a legal value-add — not just a UX choice