The Library of Congress holds a small color advertising card printed by the Burrow-Giles Lithographic Company at 20 and 22 Gold Street in New York. The front shows Oscar Wilde on a sofa with a closed book in one hand and his head resting on the other, and the back lists prices for hats and bonnets. Some 85,000 copies of it were sold or offered for sale, the photographer who made the picture sued, and that is how the Supreme Court came to decide whether a photograph can have an author at all.
The Sitting on Union Square
Napoleon Sarony was born in Quebec in 1821, spent his early career in lithography, and by the early 1880s ran the best-known celebrity portrait studio in New York from a building on Union Square. He photographed Mark Twain, Walt Whitman, Lillian Russell, and Sarah Bernhardt. He was also among the first American photographers to treat the famous face as a business input, paying sitters for the right to make and sell prints of them.
Wilde arrived in January 1882, 27 years old, with no novel and no produced play behind him, on a lecture tour tied to the American staging of Gilbert and Sullivan's Patience. Sarony worked him through at least 27 setups: fur-lined overcoat, evening dress, and the velvet jacket with knee breeches and silk stockings that became the image of Wilde for the rest of his life and well past it. Each frame got a number, the set was deposited for copyright, and the prints carried the line "Copyright, 1882, by N. Sarony."
The deposit set is still in Washington, cataloged as LOT 12385, an incomplete run of numbered studio portraits that includes number 18. What the trial court later found about the arrangement between the two men is thinner than the anecdotes suggest: the record says only that Sarony became the author of the picture "under an agreement with Oscar Wilde." Accounts differ on whether money changed hands and in which direction, and the case file does not settle it.
None of that was unusual for the period. Cabinet cards of stage and literary celebrities were a mass-market product, and unlicensed reprints of them were a standing tax on the studios that made them. What was unusual was that Sarony had the money and the temperament to litigate.
The Machine Argument
Congress had already written photographs into copyright. The Act of March 3, 1865, extended protection to photographs and their negatives on the same terms as prints and engravings, and Abraham Lincoln signed it six weeks before he was killed. By 1882, that provision had been folded into section 4952 of the Revised Statutes, which listed photographs alongside books, maps, charts, engravings, paintings, drawings, and statuary.
Burrow-Giles could not argue that the statute failed to cover the picture, so it argued that Congress had no power to pass the statute in the first place. The Constitution authorizes protection for the writings of authors. A photograph is not a writing, the company said, and the man who trips the shutter is not an author. Justice Samuel Miller set out the argument at length before answering it: the chemicals and the machinery were fair subjects for a patent, but everything after that was "merely mechanical, with no place for novelty, invention or originality." The camera work itself, on that account, was nothing but hand-operation, copying a thing that already existed onto a plate, with "the accuracy of this representation being its highest merit."
Strip out the 19th-century vocabulary, and that is the argument now made about generative image tools, pointed in the opposite direction. A machine did the work. The person only ran it. Whatever came out cannot belong to anybody.
The Court Answered With a List of Choices
Sarony had already won below. A judge sitting without a jury in the federal circuit court in New York gave him $600 for the plates and the 85,000 copies, plus $10 for the copies still in the defendant's hands, as statutory penalties. The company took it up on a writ of error, the case was submitted on December 13, 1883, and it was decided on March 17, 1884, with no justice dissenting. Miller could have written a broad rule about photography as a medium. He wrote about one photograph instead, and he did it by quoting the trial court's third finding of fact almost in full.
Sarony, the finding said, made the picture "entirely from his own original mental conception, to which he gave visible form by posing the said Oscar Wilde in front of the camera, selecting and arranging the costume, draperies, and other various accessories in said photograph, arranging the subject so as to present graceful outlines, arranging and disposing the light and shade, suggesting and evoking the desired expression." On the strength of that list, the Court held the picture to be "an original work of art, the product of plaintiff's intellectual invention."
Read it again as a working photographer, and it stops sounding like law. It is a shot list. Where the subject sits, what gets draped behind him, which way the key falls, what you say to him to get the face you want. The case turned on a description of ordinary studio craft, the kind of decision-making that any competent portrait shooter still makes in the first ten minutes of a session, and the Court treated that sequence of choices as the authorship. If you want the modern version of the same skill set, lighting a portrait sitter has not changed nearly as much as the legal stakes attached to it.
Miller also defined the operative word. An author is "he to whom anything owes its origin; originator; maker; one who completes a work of science or literature." Copyright, he wrote, is "the exclusive right of a man to the production of his own genius or intellect." The constitutional holding follows from those definitions, and it is narrower than the case's reputation: the Constitution reaches copyright in photographs, Miller wrote, "so far as they are representatives of original intellectual conceptions of the author."
What the Court Refused to Decide
The shorthand version is that Burrow-Giles v. Sarony settled that photographs are copyrightable. That is how it gets taught, how it gets cited in briefs, and how it appears in most timelines of American copyright. The shorthand is not crazy. The Court did uphold a photographic copyright against a constitutional attack, and no later court has seriously suggested that photographs as a class fall outside the statute.
But the opinion refused to say that. Immediately after laying out the mechanical-reproduction argument, Miller wrote: "This may be true in regard to the ordinary production of a photograph, and, further, that in such case a copyright is no protection. On the question as thus stated we decide nothing."
That is the Court conceding that a photograph made by merely working the machine might have no protection at all, and then declining to rule on it. The syllabus makes the same point in procedural terms: "Whether a photograph is a mere mechanical reproduction or an original work of art is a question to be determined by proof of the facts of originality, of intellectual production, and of thought and conception on the part of the author," and when someone contests the copyright, that proof has to be produced. Sarony won because he could prove what he did. The choiceless photograph was left standing outside the courthouse.
What happened next was not an answer to that question. It was a steady lowering of the bar for everything else. Lower courts began reading the originality requirement loosely within a decade, and in Falk v. Brett Lithographing Co. in 1891, a New York federal court upheld a studio photograph's copyright on a noticeably lighter showing of arrangement. The bar dropped much further in 1903, when the Supreme Court decided Bleistein v. Donaldson Lithographing Co., a fight over advertising posters for a circus. Justice Oliver Wendell Holmes wrote that "the copy is the personal reaction of an individual upon nature," that "personality always contains something unique," and that "a very modest grade of art has in it something irreducible, which is one man's alone." After that, almost any photograph with a person behind it cleared the standard. What Bleistein did not do was answer the question Miller reserved. It was a case about circus posters, and its reasoning runs on the premise that a human reacted to something; it says nothing about a picture with no choices in it at all. That question was never closed, and it has resurfaced whenever somebody has tried to make a photograph with the choices deliberately taken out. In 1999, in Bridgeman Art Library v. Corel, a federal district court in New York held that exact transparencies of public-domain paintings were not original enough to copyright. Judge Lewis Kaplan went back to Burrow-Giles to explain why, noting that the Supreme Court there had "declined to decide whether 'the ordinary production of a photograph' invariably satisfies the originality requirement," and then answered the piece of it in front of him: "slavish copying," he wrote, "although doubtless requiring technical skill and effort, does not qualify." What generative image tools changed is not the question. It is how much money now sits on the answer.
The Same Argument, Now About Prompts
In January 2025, the US Copyright Office published Part 2 of its report on copyright and artificial intelligence, on the question of what AI output can be registered. It reaches Burrow-Giles on page nine, quotes Sarony's posing, costume, and expression choices, and draws the rule out of them: "the use of a machine as a tool does not negate copyright protection, but the resulting work is copyrightable only if it contains sufficient human-authored expressive elements."
Applied to text-to-image tools, the Office's conclusion is blunt. "[G]iven current generally available technology," it wrote, "prompts alone do not provide sufficient human control to make users of an AI system the authors of the output." Prompts convey ideas, and ideas have never been protectable. Grinding through revisions does not help either, because by resubmitting prompts the user is "re-rolling" the dice, generating more candidates to choose from without gaining control over how any of them are made.
The report deals with the photographic objection directly, because commenters raised it. Several argued that a camera left running on a tripod is no more controlled than a prompt, and that a wildlife frame nobody was present for is protected anyway. The Office's answer is that a nature photograph's protection rests primarily on the angle, location, speed, and exposure the photographer chose. The human is still responsible for executing the expressive decisions. That is Sarony's finding of fact restated for a tripod.
The registration record tracks that line closely, and it cuts both ways for photographers. In March 2023, the Office registered an image built on a human-drawn input, annotating the certificate to limit the claim to "unaltered human pictorial authorship that is clearly perceptible in the deposit and separable from the non-human expression that is excluded from the claim." In December 2023, its Review Board affirmed a refusal for an image made by feeding a photographer's own photograph into an AI painting app with an instruction to render it in the style of van Gogh's The Starry Night. The output did not contain "sufficient human authorship necessary to sustain a claim to copyright," the Board found, because the applicant had too little control over what the system did with the file.
The courts have not disturbed any of that. The D.C. Circuit held in March 2025 that human authorship is required by statute, in the case of an image its applicant said was generated autonomously by software, and the Supreme Court declined to hear the appeal on March 2, 2026. A closer test is still pending in the District of Colorado, where Jason Allen is challenging the refusal to register Théâtre D'opéra Spatial, the Midjourney image that won a Colorado State Fair art prize. Allen revised and re-entered prompts at least 624 times, cleaned up the result in Photoshop, and enlarged it with Gigapixel AI. The government's position is narrower than it sounds. Its brief argues that Allen did not control the expressive visual elements Midjourney generated and had to disclaim them. It concedes the rest. The Office told him that "the final artwork does show that Mr. Allen added new creative authorship to the Midjourney-generated image," and the government's filing states that he "likely could have successfully registered the Work had he simply agreed to limit his claim by disclaiming the underlying material that he did not author." The fight is over the disclaimer, not over whether he contributed anything. The court had not ruled as of this writing.
What Would Survive Today
The parallel has limits. Sarony was not arguing that his camera deserved rights, and no photographer today is arguing that a diffusion model does. What repeats is the evidentiary shape of the fight. In both centuries, the challenged work came out of a machine, the defense was that the machine did the work, and the answer turned on whether a human being could point to expression that was actually theirs.
Sarony could. The 1884 opinion is not really about photography as a medium; it is about proof. The trial court wrote down where Wilde sat, what was hung behind him, how the light was placed, and what was said to him to produce that expression, and the Supreme Court read that list and found an author in it. Note who wrote it. The trial court did, at trial, not Sarony during the sitting. Nothing required him to keep a log while he worked, and the timing of the contribution is not the test either. The Copyright Office's current position is that human authorship "may be added in the final product," and that creative modifications of an AI output can be protected in their own right.
No rule requires you to keep a record, and nobody will refuse you a copyright for lacking one. But these fights turn on proof, and proof is easier to assemble on the day than three years later. If you ever have to defend authorship in a frame that a client, a stock platform, or an AI-assisted workflow puts in question, the document that does it will look like Sarony's third finding of fact. What you moved, what you built, where you put the light, what you asked the subject to do. That is worth writing down on the day, while you still remember it.
Lead image by Napoleon Sarony, public domain. Source.

5 hours ago
19







English (US) ·