If you’ve used any kind of artificial intelligence tool to help create art, photography, music, or writing, 2026 is the year the legal ground shifted under your feet. A years-long dispute over a single AI-assisted image called SURYAST has become the defining test case for AI authorship copyright 2026 questions, and its outcome will shape how the U.S. Copyright Office treats every AI-assisted work that comes after it. Whether you’re a photographer who ran an image through a style filter, a novelist who used AI for brainstorming, or a small business owner generating marketing graphics, the SURYAST saga touches your work more directly than you might think. In this article, we’ll walk through what the SURYAST case actually says, why “human authorship” has become the make-or-break legal standard, and — most importantly — what you can do right now to protect your claim to a work, regardless of how the courts eventually rule.

Understanding AI Authorship Copyright 2026: The SURYAST Case Explained

SURYAST is a two-dimensional image created by intellectual property lawyer Ankit Sahni. He started with an original photograph he took himself, then ran it through an AI painting application called RAGHAV, applying the visual style of Vincent van Gogh’s “The Starry Night” at a strength value he selected. The Respondent, Ankit Sahni, generated the output at the center of this case, SURYAST, using an AI application called RAGHAV, which then processed his inputs to generate the final output. That sounds simple enough — but the question of who, or what, actually “authored” the resulting image has now been fought over in at least three countries and multiple courts.

The U.S. Refusals

Sahni first submitted SURYAST to the U.S. Copyright Office in March 2022. The Copyright Office refused the application in June 2022, saying the work “lacks the human authorship necessary to support a copyright claim” and that because the work “was in part generated by a computer program,” the office was “unable to register” the claim. On reconsideration, the Copyright Office refused again in April 2023, calling the work a “derivative work” that did “not exhibit the requisite human creativity needed to support a claim to copyright in a derivative work.” The Copyright Review Board issued its final word in December 2023, finding the work was “not the product of human authorship” because “the expressive elements of pictorial authorship were not provided by Mr. Sahni.” Notably, the Board left the door open for copyright registration of a work authored by a human with assistance of AI, but concluded that Sahni’s input was insufficient to rise to that level — acknowledging he conceived of the idea while finding RAGHAV was the author of the expression of that idea. The Board summed up its concern bluntly: Sahni “exerted insufficient control over RAGHAV’s creation of the work.”

The Case Isn’t Over

This didn’t end the fight. In May 2026, a new lawsuit escalated the dispute into federal court. A California company sued the U.S. Copyright Office after the agency refused to register Suryast, with the lawsuit filed May 8 in U.S. District Court for the Central District of California by an entity assigned Sahni’s rights in the disputed work. Meanwhile, the story looks different abroad: India’s copyright office registered Suryast in 2020 with Sahni and the software RAGHAV listed as co-authors, after initially rejecting an application that listed the app as the sole author. Canada is now weighing in too — the advocacy group CIPPIC filed a case seeking to correct the Canadian registry, and a similar proceeding is before the Federal Court of Canada in CIPPIC v. Sahni, which seeks to expunge or amend a Canadian copyright registration naming an AI system as a co-author, potentially giving a Canadian court its first opportunity to address these authorship and evidentiary questions. You can read CIPPIC’s own summary of its position on the CIPPIC v. Sahni case page.

Why “Document Human Authorship AI” Is the New Legal Battleground

SURYAST isn’t an isolated dispute — it’s part of a broader pattern. Copyright offices are potentially refusing registrations where the human contribution to an AI-assisted work cannot be clearly demonstrated