As AI creeps into their practice, creative industry professionals are being forced to take a stance, develop policies, and navigate new client expectations. “You are seeing a divergence in the marketplace,” she says. “Those who say it’s all about cost, and those who say it’s all about human creativity.”
Changing expectations
In the middle of it all are the agents representing photographers, who must navigate a fine balance between meeting the needs of their commercial clients, and representing their photographers’ interests. While work is still coming in, one of the first major challenges agents have experienced in photography commissioning since AI entered the equation is the way that it is changing client expectations. Agents are receiving highly specific AI-generated mock-ups, known as scamps, that not only leave less room for the artist’s authorship, but are also raising expectations around what’s possible.
“At the moment, the most common way AI is appearing for us is coming from client’s internal use; things like pre-visuals, briefs, creative mock-ups, and storyboards that we come into contact with as projects come in,” says Hati Gould, an agent at East Photographic. “Clients are presenting mock-ups that are often very close to what they want the final outcome to be.”
AI mock-ups are hyper-specific and hyper-realistic in a way that sketches and moodboards never were, so clients arrive with a fixed vision rather than a direction. They’ve often been signed off internally, locking in expectations. And because they look like finished images rather than rough concepts, the gap between the brief and what’s practically possible is harder to explain.
Laura Dawes, director at Webber — an international agency representing photographers, directors, stylists, and set designers — says that the AI mock-ups one client had weren’t possible to produce in the conditions of the shoot. In response, Dawes says Webber has updated its terms of contracts to reflect new scenarios: “Any kind of scamps [mock-ups], pre-production briefings, or approvals that use AI have to be signed off or approved by us, just to ensure that they can deliver what the client has asked for.”
Post-production in a post-AI world
Elsewhere, AI is appearing in new scenarios in post-production, too. Charlotte Long, head of photography at Academy Films, describes a fashion shoot where a photographer had shot stills for a client, but by the time the brand shared it on social media the images had become motion assets. “It was alarming at first,” she says, “but also very intriguing. And to be honest, it was really impressive with how they’d done it.” However, factoring in such usage earlier into a job might have had a different creative outcome. “If the photographer knew they were going to be delivering videos, they might have been lit in a different way,” she adds.
Where some clients have explored fully AI-generated campaigns, Long finds that the work which begins with a photographer’s original image — even if AI is used somewhere in the production process — is both logistically and legally cleaner. There’s an original file to edit from, and the photographer owns the IP. “It’s much easier to navigate the usage if the photographer already owns the usage,” she says. Although, if real people and models are involved, usage becomes trickier when it comes to negotiating AI usage terms that “some model agencies don’t agree to as well”.
Meanwhile, some photographers and agents are trying to shield their work from being fed into AI once it leaves their hands. Contracts are adjusting to control that kind of usage, and though it is hard to monitor, emerging services such as Glaze and Nightshade are claiming to help protect creative works by affecting how AI services can read them.
