← All perspectives
Photography & technology · Essay

What are we asking a photograph to prove?

AI can speed up an edit, invent a scene, or replace a commission. Those changes ask very different things of photographers and the people hiring them.

UpCrafted editorial · 11 September 2026 · 6 min read
Illustration of a camera, softbox, and studio backdrop

A photographer uses software to sort a long shoot. A designer generates a fictional landscape. A brand replaces a planned portrait session with a synthetic person. Each might be described as using AI in photography, but the phrase conceals more than it explains.

The first changes a workflow around a captured event. The second creates an image whose subject may never have existed. The third changes who gets hired, what the audience sees, and the relationship between a brand and the people it depicts. A useful conversation has to keep those distinctions visible.

This matters because photography is asked to do several jobs. It can create atmosphere, demonstrate a product, document an event, describe a person, or help us notice something unfamiliar. An invented image may serve some briefs. Other briefs depend on someone having been there. The question for a creative team begins with the promise the image makes.

Before choosing how to make an image, decide what the audience is entitled to believe about it.

Relief in the editing room is a real part of the story

Aftershoot’s account of its 2025 photography research describes more than 1,000 working photographers, with a focus on early AI adopters. It reports that 81% said their work-life balance improved and 64% said clients did not notice AI-assisted editing. These are photographers’ reports about their experience and client reactions. They are not results from a blinded audience test or a representative survey of every photographer. Aftershoot also sells the tools being discussed.

Even with those limits, the result points to a practical distinction. A tool that reduces repetitive sorting can be valuable without replacing the shoot. Time returned to a photographer might support more careful selections, better preparation, or simply a reasonable evening away from the desk.

The creative responsibility remains in the review. A technically imperfect frame can contain the gesture that makes the sequence work. A consistent automated adjustment can still be wrong for a particular face, light, or atmosphere. The useful workflow makes those judgments easier to exercise, rather than quietly removing the opportunity to exercise them.

Source: Aftershoot · discussion of its 2025 survey of AI-adopting photographers

A faster workflow does not guarantee a healthier market

The Association of Photographers offered a sharply different perspective in its January 2026 response to generative AI, describing member concerns about lost commissions, licensing, and income. The AOP is an organization advocating for photographers, and its statement should be read in that context. Its public release does not provide enough methodological detail to turn its financial figures into an estimate for the whole profession.

The two accounts need not cancel each other out. A photographer may spend less time editing each assignment while finding fewer assignments available. Efficiency within a job and demand for that job are separate questions. A single cheerful productivity number cannot settle both.

For a creative director, the issue becomes concrete when a budget is allocated. What would the commission discover that a generated image would not? Is the brief asking for a convenient visual symbol, or a meaningful encounter with a person, product, or place? Be honest about the answer. Asking a photographer to defend every commission through vague claims about human authenticity is no better than assuming the cheapest plausible image is always adequate.

Source: Association of Photographers · statement on creative livelihoods, January 2026

Plausibility and evidence have different standards

World Press Photo’s announcement of its 2025 winners explicitly prohibited AI-generated images and generative fill. That rule belongs to a documentary and photojournalistic contest; it is not a universal rule for illustration or advertising. It makes the purpose of the work unusually clear: viewers must be able to engage with photographs as accounts of the world.

A brand team can ask a related question without pretending that every campaign is journalism. Does the image imply that this employee works here, that this meal was served, or that this product performed in the scene shown? If the answer matters to the message, the production method matters too. A fictional scene needs an appropriate framing; it should not casually borrow the authority of a documented event.

Real photographs also require judgment. Framing, selection, captioning, and editing influence what viewers understand. Commissioning a shoot is the beginning of responsibility for an image, not a guarantee that every later use will be honest. The brief and the edit should consider what is outside the frame as carefully as what looks good inside it.

Source: World Press Photo · 2025 contest announcement and AI policy

A history of the file helps, but it cannot do all the work

The C2PA standard supports information about the origin and editing history of digital content. Its explainer distinguishes that provenance from a judgment that the depicted content is factually true. A valid history can help establish who made assertions about a file and whether those records have been altered. It cannot establish the honesty of every scene or caption. Missing credentials are not, by themselves, proof that an image is fake.

For a team commissioning photography, useful habits start before the final export. Keep the agreed brief, original files, selection notes, and an account of material edits. Decide what context needs to travel with the image and who is responsible for preserving it. When a picture is handed to another team, the explanation should not depend on a person remembering a conversation six months later.

Transparency works best when it is understandable. A client does not necessarily need a list of every software operation. They do need to know when a change affects the meaning of the image, the identity of the subject, or what the picture appears to document. Explain those decisions in language that survives beyond the production team.

Source: C2PA · Content Credentials explainer, version 2.4

Write a brief that respects the image and the maker

Start with what must be real. That might be the person, the product, the location, or an event that cannot be repeated. Then agree which kinds of intervention fit the purpose: routine adjustments, extensive retouching, compositing, or generated elements. Naming those choices early gives the photographer something more useful than a late request to make it feel authentic.

Discuss intended uses, approvals, and future reuse before production begins. If a team wants to upload commissioned material to an AI service or use it in a new synthetic workflow, that deserves a separate conversation with the people responsible for the work. Treat consent as an actual agreement rather than an assumption attached to possession of a file.

For photographers, a portfolio can explain more than a recognizable look. Show the observation, access, timing, direction, and problem-solving behind a series. Make visible what the assignment required and why the result mattered. This cannot remove market pressure, but it gives a commissioner a more precise account of what they are buying.

For designers, stay curious about the photograph before turning it into an asset. Ask about the moment, the subject, and the decisions behind the frame. That curiosity can lead to a better crop, a more accurate caption, or a different brief entirely. As image-making becomes easier to initiate, knowing what an image is for becomes a more consequential part of the job.