If you model for a living, your face is your inventory. A new complaint out of Sydney asks what happens when a brand keeps your face on file and quietly edits it. Nigerian-Australian model Elii Emeghebo has filed a complaint against menswear retailer Peter Jackson, alleging the company used AI to lighten his skin, narrow his features, and change his eye color in ads built from his own photo shoot. He says he found the altered image by accident, walking past the brand's shop.
The claim
According to the complaint, filed with the Australian Human Rights Commission, Emeghebo previously worked for Peter Jackson and later noticed an ad that used his likeness in visibly altered form. He described the changes directly: "My nose was reshaped, my skin tone and my eye colour was significantly lightened and there's some reshaping around my eyebrows and my eye shape to be more Eurocentric and a lot less black," he told reporters. The report notes the lawsuit is not the first to allege that companies are using AI to alter the likeness of certain models, pointing to the parallel case of model Francheska Pujols, who sued Rainbow Shops in New York after the retailer allegedly generated new, uncommissioned images of her from an existing shoot.
Why it matters
Most of the AI-likeness disputes we have covered this year sound in right of publicity or deepfake statutes: a model's face used without consent, in a pose or context she never agreed to. Those claims focus on control and consent. Emeghebo's complaint asks a different question, whether an AI edit that erases someone's racial features is discrimination, not just a contract or consent violation. That is why it went to a human rights body rather than a court seeking damages under a publicity statute. It is a novel theory, and untested. Independent models of color booking AI-heavy campaigns should understand that the existing patchwork of protections, built mostly around consent and likeness rights, was not designed with algorithmic bias in mind. As one recent overview of the space put it, because there is no comprehensive federal AI statute in the US, brands operate within a patchwork of legacy doctrines and emerging AI-specific laws, and the same tools that let brands 'do more with less' also make it easy to reuse a model's likeness in ways they never agreed to. Nothing in that patchwork was written with skin-tone alteration in mind. New York's approach shows one direction regulators are already moving. On June 19, New York imposed new consent requirements for the use of generative artificial intelligence tools that substantially replicate or replace a model's appearance or performance. That law addresses whether an edit happened at all. It does not, on its face, address what kind of edit happened, or whether that edit reflects bias baked into the underlying AI system or its prompts.
What to watch
Watch whether the Australian Human Rights Commission accepts this as a discrimination matter rather than routing it back toward a straightforward likeness or contract dispute. Watch, too, whether any U.S. plaintiff brings a similar claim, since American civil rights statutes were not built around AI image alteration and the fit is uncertain. For independent designers and small brands using AI tools to touch up campaign photography, the practical lesson is not subtle. If you commission a real model, keep her actual features. If your AI workflow lightens skin, narrows noses, or otherwise moves a subject toward a different racial appearance, you are not making a minor edit. You are creating a new person, and you may be doing it in a way that a regulator, a court, or your own model, will not accept as harmless retouching.