Who Owns AI-Generated Content? A Global Snapshot for 2026
Why the answer differs depending on whether you're in the US, UK, India, or EU, and what that means for businesses using AI tools commercially
The Core Problem
Copyright law was built around the assumption that a human being creates the work. Generative AI breaks that assumption, and every major jurisdiction is now wrestling with the same basic question in slightly different ways: if a tool produces an image, a piece of music, or a block of text with minimal human input, who, if anyone, owns it.
The honest answer in 2026 is that the law is settled on one point and unsettled on almost everything else. Purely AI-generated output, with no meaningful human creative contribution, is not protected by copyright anywhere that has ruled on the question. What counts as "meaningful human contribution" is where jurisdictions start to diverge, and where the real legal risk for businesses sits.
The United States: A Clear, Hard Line
The US has the most settled position of any major jurisdiction, and it's a strict one. In March 2026, the Supreme Court declined to hear Thaler v. Perlmutter, the case brought by computer scientist Stephen Thaler seeking copyright protection for an artwork his AI system DABUS generated autonomously. By refusing to take the case, the Court left intact the D.C. Circuit's ruling that copyright requires a human author, full stop.
The reasoning matters for how the rule gets applied going forward. The appellate court grounded its decision in the statutory language itself, noting that the Copyright Act's references to a work's duration extending through an author's life, or passing to their heirs, presuppose a human being with a finite lifespan. A machine simply doesn't fit that structure.
In practice, the US Copyright Office now requires applicants to disclose how much of a submitted work was AI-generated versus human-authored. Purely AI-generated content gets rejected outright. Work where a human made meaningful creative choices, selecting, arranging, editing, or substantially modifying AI output, can still qualify, but each registration is being assessed case by case, and simply writing a detailed prompt is not, on its own, enough to count as authorship.
The United Kingdom: The Outlier
The UK took a different path decades before generative AI existed. Its Copyright, Designs and Patents Act 1988 includes a specific provision for "computer-generated works," works produced by a computer in circumstances where there's no human author, and assigns authorship to the person who made the arrangements necessary for the work to be created.
This sounds like it should resolve the AI question neatly, but it doesn't. The provision was written in 1988 with far simpler computer-assisted creation in mind, and nobody has definitively tested whether it can stretch to cover modern generative AI, or whether such output can even meet the originality threshold the rest of UK copyright law requires. The UK remains the jurisdiction most structurally open to recognising AI-assisted authorship, but the open question of how that openness applies to today's tools is still just that: open.
India: An Intermediate, Unsettled Position
India's Copyright Act, 1957 sits somewhere between the US and UK approaches, and the gap is currently being filled by Copyright Office practice rather than clear statute. Section 2(d)(vi), added in 1994, defines the author of a computer-generated work as the person who causes the work to be created, language that closely mirrors the UK's approach.
In practice, though, India's Copyright Office has applied this narrowly. The most cited example is the RAGHAV case, where artist Ankit Sahni initially secured registration listing himself and his AI tool as co-authors of an artwork. The Copyright Office later issued a withdrawal notice, taking the position that the 1957 Act requires authorship to rest with a natural person and that AI cannot hold co-authorship status. The application technically remains listed as registered while the underlying question stays unresolved, but the direction of Copyright Office practice is clearly toward requiring a human author, much like the American approach, despite the more flexible statutory language.
For businesses operating in India, the practical guidance from most practitioners right now is straightforward: treat AI as an assistive tool, document the human creative decisions made during the process, and avoid relying on AI tools as the named or co-named author in any registration filing.
The EU: No Special Provision, Same Underlying Requirement
The EU hasn't legislated specifically on AI authorship, but its general originality standard, that a protected work must reflect the author's own intellectual creation, functions as a similar human-authorship gatekeeper. A work generated without meaningful human creative choices is unlikely to satisfy that standard regardless of which member state's courts hear the dispute. The bigger live battleground in Europe right now is less about who owns AI output and more about whether AI companies infringed copyright by training models on protected works without permission, an issue regulators and courts across the bloc are actively grappling with.
What This Means for Businesses Using AI Tools
A few practical conclusions hold across all four jurisdictions, even though the legal tests differ in their details.
Simple prompt-to-output generation, where a user types a request and takes the first result with no further editing, is unlikely to be protectable anywhere. If brand or commercial value depends on owning the rights to AI-assisted content, document the human creative process: iterations, selections, edits, and original elements added on top of AI output all strengthen the authorship claim.
Whether you can commercially sell or license AI-generated content is a separate question from whether you own the copyright in it, and is often governed entirely by the platform's terms of service rather than copyright law itself. Read those terms carefully before building a business model around a particular AI tool's output.
Finally, the law in every jurisdiction discussed here is still actively developing. Litigation over AI training data, ongoing in the US and gaining momentum in the EU, may reshape the landscape well before authorship rules themselves are fully settled. Businesses making significant investments in AI-generated content pipelines should build in flexibility rather than assuming today's rules will hold for the next several years.
This Blog is for general informational purposes and does not constitute legal advice. For guidance specific to your content, IP, or AI licensing strategy, please contact our team.