Driver FixRecommendedSound, Wi-Fi or graphics acting up? Check drivers firstFind missing or outdated drivers fast.Check DriversOctober DealsAmazon USOctober deal check: compare before you payAmazon US: current deals, useful picks and tech finds.Check DealsPC HealthRecommendedCrashes, freezes, slowdowns? Check your PC nowSpot repairable issues before they interrupt work.Check PC×
Skip to content
DocumentaryTube
AI-generated imagery

Vogue Just Used AI-Generated Imagery in a Nightmare-Fuel Cover Shoot

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Yes—but Bella Hadid was photographed for real. Vogue Italia’s May 2023 Ars Artificialis issue combined studio photographs of Hadid and her fashion styling with AI-generated scenery and image extensions made with OpenAI’s DALL-E 2. The result looked uncanny because the subject was physically present while the surrounding world was invented, stretched and composited by people and software.

What Vogue Italia actually published

The project appeared in Vogue Italia’s May 2023 issue, Ars Artificialis, on pages 138–155. Vogue’s description called the pictures “real shots on imaginary backgrounds created by the artificial intelligence program DALL-E.” The central subject, Bella Hadid, was not an AI avatar: she was photographed in a studio wearing Imruh Asha’s styling.

The credited team was photographer Carlijn Jacobs, stylist Imruh Asha and AI artist Chad Nelson, working with Hadid. Vogue Italia published its account on 27 April 2023. A Futurism report dated 29 April 2023 characterized the project as an early fashion-magazine use of AI-generated imagery.

How the hybrid images were made

1. The team photographed the real subject

Jacobs and Asha developed the visual concepts and shot Hadid in a studio. Her pose, clothing, skin, lighting and shadows came from a physical photo session rather than a text prompt.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

2. DALL-E supplied the imaginary world

The team generated backgrounds and extensions with DALL-E 2. Instead of accepting the first result, Jacobs moved from loose prompts to detailed keyword descriptions. Nelson helped refine prompts and increase resolution.

3. Reference images guided the generations

They supplied reference images, including outtakes and material showing Hadid’s lighting, shadows, texture and clothing. Those references helped the generated environment fit the photographed figure.

4. Humans selected, composited and retouched

The outputs were judged and revised repeatedly, then combined and finished in Photoshop. Vogue says the method had to be revised several times. Producing an image that did not look artificially assembled was more complex and time-consuming than the team initially expected.

Why the pictures feel like “nightmare fuel”

The unease comes from the collision of photographic evidence and synthetic space. Hadid’s body carries the visual cues of a real camera—specific fabric, texture and directional light—while the surrounding architecture, landscape or objects can obey different rules. Small errors in perspective, anatomy, scale or shadow make the scene feel almost credible, then suddenly wrong.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

That tension was part of the appeal of the concept. The images were not presented as ordinary documentary photographs; they deliberately treated fashion photography as a place where a real person could inhabit an impossible setting.

What still required human judgment

This was not a one-command generation. Jacobs described collaborating with AI as being like a curator assisting a machine’s creative process. Nelson compared working with DALL-E to “a visual slot machine,” a phrase that captures the unpredictable variation between prompts and outputs.

  • Concept: Jacobs and Asha decided what the image should communicate.
  • Direction and styling: Hadid’s pose, clothing and studio lighting were planned and photographed by people.
  • Prompt design: The team converted visual ideas into increasingly precise descriptions.
  • Selection: Humans rejected unusable generations and chose the elements that supported the composition.
  • Finishing: Photoshop compositing and retouching corrected joins, scale and visual continuity.

Asha said AI was convenient for mocking up ideas quickly but that she would not use it to create the final product. Jacobs likewise said the experience augmented and complemented her photography skills rather than replacing them.

AI-assisted fashion imagery versus a conventional shoot

Question Hybrid method used by Vogue Conventional production
What is photographed? Hadid and the styling were photographed in a real studio; backgrounds and extensions were generated. Subject and physical or digitally designed setting are produced through the normal photographic and art-direction workflow.
Control of pose and lighting High control over the photographed subject, with additional work needed to match the generated environment. Control is established through the location, set, lighting and post-production plan.
Iteration Prompt changes and many candidate generations, followed by human selection and Photoshop work. Changes generally require new set, location, prop, lighting or post-production decisions.
Logistics Can avoid building or traveling to every fantastical setting, but introduces integration and retouching work. May require construction, travel, permits, crews and physical materials for an unusual setting.
Rights and consent Includes the rights and consent issues of the photo shoot plus unresolved questions about AI training data and artist compensation. Rights normally center on the commissioned photographs, people, locations, designs and licensed assets.
Creative control Human direction remains essential, while the model supplies unpredictable visual options. Human direction controls the concept and execution without a generative model’s variations.

What went wrong during production

Vogue’s account identifies perspective, integration, anatomy and production time as practical problems. Generated elements did not automatically share the camera’s viewpoint or the subject’s lighting. The team had to iterate prompts, provide better references and repair mismatches in post-production.

Free tools Windows power users keep installed

One-click scans. No signup required.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Jacobs estimated that making images without an artificial appearance could require up to five times more work than traditional shooting methods. That is a qualitative estimate from the photographer, not an independent time-and-motion study, and it should not be treated as a general industry benchmark.

Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Support on Ko-Fi

The copyright and labor questions

The visual experiment also exposed unresolved business and ethical issues. Futurism’s 29 April 2023 report raised questions about the data used to train commercial image generators, how artists are compensated and whether AI imagery substitutes for human labor.

Those questions are separate from whether Hadid herself was photographed. A real model and photographer can participate in a shoot while the generated background still draws on a system whose training sources and compensation practices may be contested. The project therefore demonstrates augmentation, but it does not settle the larger argument over ownership, consent or employment.

Was the cover photographed or generated?

It was both, in different parts. Hadid, her clothes and the initial studio lighting were photographed. DALL-E 2 generated the imaginary backgrounds and extensions, and people composited and retouched the final pages. Calling it a wholly AI-generated cover is inaccurate; calling it an ordinary photograph is also incomplete.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Why this 2023 experiment still matters

Vogue Italia’s shoot is useful as a case study because it makes the division of labor visible. The model did not replace the camera, the stylist or the photographer. Instead, a generative system became an unpredictable source of scenery that required curation, technical correction and substantial finishing.

For viewers, the key question is not simply whether an image is “real.” It is which parts were captured, which parts were generated, who directed the choices and whose work or data supported the result. Vogue’s pages answer the first question clearly: real subject, imaginary surroundings.

Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.

Leave a Reply

Your email address will not be published. Required fields are marked *

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Read next

Recommended PC Tool
Recommended PC Tool
Outdated Drivers Are Slowing You DownFree scan - exact matches
Windows Errors? Fix Them Before They SpreadFree repair scan

Two free Windows tools

One Free Minute Could Fix That PC

Before you go - each of these free tools takes about a minute and tackles what quietly slows a Windows PC down.

Special offer. View Outbyte info, uninstall instructions, EULA, and Privacy Policy.