If this is manipulated, then the floodgates are open.
For nineteen years, Petra Urban has rendered Dutch courtroom proceedings in pencil — neutral, faithful, belonging to no party. When a far-right political office fed her sketch of two Syrian brothers into an AI and posted the darkened, sharpened result as implicit propaganda, they crossed not merely a legal line but a philosophical one: the boundary between documentation and distortion, between the public record and the political weapon. The MP apologized and paid damages, but Urban's deeper concern endures — that as AI manipulation grows cheaper and easier, the very idea of neutral journalistic witness becomes something that must be actively defended, not simply assumed.
- A court artist discovered her work had been algorithmically transformed — features sharpened, expressions hardened — and deployed as political imagery without her knowledge or consent.
- The violation struck at three levels simultaneously: copyright theft, political weaponization of independent work, and the unsettling novelty of a machine being used to reshape human documentation into propaganda.
- Urban's union moved formally against the MP responsible, invoking Dutch moral rights law, which protects creators from distortions that damage their professional reputation — a legal framework rarely tested against AI manipulation.
- The MP apologized and paid undisclosed damages, but the settlement resolved only the individual case, leaving the broader vulnerability of visual journalists entirely intact.
- Urban's warning — that manipulated journalistic work opens the floodgates — lands with particular weight given the MP had prior accusations of AI image alteration for campaign purposes.
- The case now stands as an early marker of a threshold: not whether AI can be used to distort neutral documentation, but whether creators retain any meaningful expectation that their work will remain true.
Petra Urban has spent nineteen years as a Dutch court artist, translating legal proceedings into pencil sketches — work she understood as neutral documentation, part of the public record but belonging to no political cause. When she sketched two Syrian brothers on trial for their sister's murder, she had no reason to expect what came next.
She found it on social media. The PVV, Geert Wilders' far-right party, had taken her drawing and run it through AI. The brothers' features had been sharpened, their expressions darkened — made to look harder and more threatening than her original depicted. The altered image was posted by the party's Noord-Brabant regional office. No permission had been sought. No payment made. No warning given.
Urban identified three distinct violations: unauthorized use of her work, its weaponization by a political faction — striking at her identity as an independent professional — and most unsettling, the algorithmic nature of the distortion itself. A machine had been enlisted to turn her art into propaganda.
Dutch law grants creators moral rights beyond standard copyright, allowing them to object when distortion damages their reputation or standing. When Urban shared the original and altered images with fellow court reporters, the story spread rapidly. Her union moved formally, demanding licensing fees and damages. The responsible MP, Maikel Boon — who had faced prior accusations of AI image manipulation — called to apologize and paid what was owed, though the amount was never disclosed.
The settlement closed the case but not Urban's concerns. She was careful to note that her objection had nothing to do with the PVV's ideology — even a party aligned with her own views would have faced the same response. The principle at stake was the integrity of neutral documentation in a democracy. 'If this is manipulated,' she said, 'then the flood gates are open.'
What her case exposed is less a single incident than a threshold. As AI manipulation becomes cheaper, more seamless, and harder to detect, the question facing every visual journalist is no longer whether such distortions can happen — but whether anyone who creates the documentary record can still expect it to remain theirs, and true.
Petra Urban has spent nineteen years drawing in Dutch courtrooms, translating the gravity of legal proceedings into pencil and paper. Last year, she sketched two Syrian brothers during their trial for their sister's murder. It was work she understood as part of a larger system—neutral documentation, the kind of visual record that belongs to the public record but not to anyone's political campaign.
She discovered the violation on social media. The Party for Freedom, the far-right PVV led by Geert Wilders, had taken her drawing and run it through artificial intelligence. The software had sharpened the brothers' features, darkened their expressions, made them look harder, meaner, more threatening than they appeared in her original work. The altered image was posted to Instagram and Facebook by the party's Noord-Brabant regional office. No one had asked her permission. No one had paid her. No one had told her it was coming.
Urban's response was precise and unsparing. She identified three separate violations. First, her work had been used without consent—a straightforward breach of copyright. Second, it had been weaponized by a political party, which struck at something deeper: her professional identity as someone who works independently, who refuses to serve any faction. Third, and most unsettling to her, the distortion itself had been algorithmic. A machine had been enlisted to reshape her art into propaganda.
Under Dutch law, creators possess what the system calls moral rights—protections that go beyond simple copyright. These rights allow an artist to object when their work is distorted in ways that could damage their reputation or professional standing. When Urban shared the original and altered images with fellow court reporters in May, the story spread quickly through the press. The violation seemed to crystallize something larger: the ease with which AI tools could now be deployed to reshape reality, and the vulnerability of anyone whose work might be useful to someone else's agenda.
Urban's union moved formally, demanding licensing fees and damages. The MP responsible, Maikel Boon, called to apologize. He paid what the court determined was owed, though the amount was never disclosed. The case closed, but Urban's concerns remained open. She spoke about what the settlement meant to her, and what it failed to address. "I hope it's clear that this is a worrying development," she said. "You need to be able to assume that journalistic work is written, drawn, photographed or filmed as neutrally as possible. If this is manipulated, then the flood gates are open."
The timing mattered. Boon had faced previous accusations of using AI to alter images for campaign purposes. Urban said she felt "no mercy" in pressing her claim. But she was also careful to separate her personal politics from her professional principle. The PVV, she said, was far from her own views—but that was almost beside the point. Even if the party had aligned with her beliefs, she would have objected. The violation wasn't about ideology. It was about the integrity of neutral documentation in a democracy, and about what happens when the tools to reshape that documentation become cheap and easy to use.
What Urban's case exposed was not a single incident but a threshold. As AI image manipulation becomes more accessible, more seamless, more difficult to detect, the question is no longer whether such things can happen. It is whether anyone who creates visual work—journalists, artists, court reporters, photographers—can maintain any reasonable expectation that their work will remain theirs, and true.
Notable Quotes
You need to be able to assume that journalistic work is written, drawn, photographed or filmed as neutrally as possible. If this is manipulated, then the flood gates are open.— Petra Urban, court artist
It compromises my neutrality.— Petra Urban, on her work being used by a political party
The Hearth Conversation Another angle on the story
Why did this particular case seem to matter so much to Urban? It's a sketch, altered and shared. Serious, yes, but why the intensity of her response?
Because it wasn't really about the sketch. It was about what the sketch represented—the idea that neutral documentation could be trusted. A court artist's job is to be invisible, to record what happens without agenda. Once that neutrality is compromised, once someone can take your work and reshape it to serve their politics, the whole system of neutral documentation breaks down.
But the MP apologized. Damages were paid. Doesn't that resolve it?
It resolves the legal claim, but not the underlying threat. Urban's point was that this will happen again, and more easily next time. The tools are getting cheaper. The detection is getting harder. One settlement doesn't stop the next person from trying.
She mentioned the PVV had done this before—manipulated images for campaigns. Why wasn't that stopped earlier?
Because there wasn't a clear legal framework yet. This case created one, at least in the Netherlands. But it's reactive, not preventive. Urban was essentially saying: we're going to keep playing catch-up unless we get serious about protecting creative work from algorithmic distortion.
What would "getting serious" look like to her?
Probably transparency requirements—forcing people to disclose when images have been AI-altered. Maybe stronger penalties. But also cultural: a recognition that this kind of manipulation is a form of dishonesty, not just a copyright violation. It's lying with someone else's work.
And if nothing changes?
Then every court artist, every photojournalist, every documentary filmmaker becomes vulnerable. Your work can be taken and reshaped to mean something you never intended. The floodgates open, as she said. There's no knowing where it ends.