Photoshop in ChatGPT: A Practical Tool for Photographers (and a Note on Transparency)
Adobe has introduced a new workflow that lets you use Photoshop from inside ChatGPT—alongside Adobe Express and Acrobat. The headline is simple: you can upload an image and request edits in plain language, then refine the results using familiar Photoshop controls. For students and working photographers, this can reduce friction in the early stages of editing and help you learn faster by turning “what I want” into “how to do it.”
This post is my photographer-to-photographer take on what’s useful here, what to watch for, and the ethics I believe matter most.
What it is (in photographer terms)
Think of it as “conversational Photoshop.” Instead of hunting through menus or remembering exact steps, you can describe your intent:
- “Lift the shadows without flattening the image.”
- “Reduce distractions in the background.”
- “Convert to black and white with luminous highlights and rich midtones.”
You can iterate: ask for one change, evaluate, then adjust. That’s important—because your style lives in restraint and refinement, not in one big “make it amazing” prompt.
The following is a sequence of prompts I gave to ChatGPT and the results. I obviously need more practice to get something I could be happy publishing. *The blur behind the caption was added to increase visibility of the caption and is not on the image itself.
The last two images were editing in Lightroom — the final image is one I would be happy with publishing – and all the finishing was done in Lightroom. (the change in the angle of the bridge is introduced by WordPress to the last image in the gallery)
Why this matters for students
If you’re learning Photoshop, the hardest part is often translating your visual goal into tools and steps. A conversational workflow can help bridge that gap.
Where it can genuinely help photographers-in-training:
-
It reduces tool overwhelm.
You can start with intent (“separate subject from background”) and learn the technique through the result. -
It speeds up first-pass corrections.
Cleanup, small distractions, and tonal balancing can go faster so you spend more time on composition, light, and story. -
It supports iterative learning.
Asking for one change at a time forces you to look critically: “Did that improve the image, or simply change it?”
Why this can be controversial in photography circles
AI-assisted editing is a hot topic—and not because photographers are anti-technology. Photography has always evolved: darkroom techniques, digital sensors, Lightroom/Photoshop, HDR, focus stacking, noise reduction, lens corrections. The friction is usually not “new tools exist,” but what those tools do and how honestly the results are represented.
Here are the main reasons this topic can trigger debate:
1) Authenticity and viewer trust
Many viewers (and many photographers) still approach a photograph as a record of something real—even when they accept interpretation in color grading or black-and-white conversion.
When AI is used to add, remove, or substantially alter elements, the image can drift from “interpretation” into “construction.” That’s not automatically wrong, but it is different. And when viewers aren’t told, trust can be damaged—especially in genres where authenticity is part of the value (photojournalism, documentary, nature, and sometimes fine art).
2) Fairness in learning and competition
In educational environments, the goal is skill-building. If AI replaces the learning process (masking, tonal control, retouching judgment), students may get “good results” without understanding how they got there.
In competitions, ethics can hinge on category rules. Some contests explicitly restrict content-aware edits, compositing, sky replacement, or generative fills. Others allow them in “open” or “creative” categories. The controversial part is not just using tools, but entering work into categories where the rules—and the expectations—don’t match the process.
3) The “slippery slope” problem
A small cleanup can become a major reconstruction faster than many people realize. Remove a distracting sign. Smooth a background. Replace a sky. Move a branch. Add a better moon. At some point, the photograph becomes a designed image.
Again, that can be artistically valid. The ethical issue is when the creator (or the platform) presents the result as if it were primarily captured rather than constructed.
4) The question of authorship
Photography is a craft of decisions: lens choice, timing, light, framing, exposure, processing choices. Some photographers feel that heavy AI-assisted edits shift authorship away from the photographer’s skill and toward a “system” that generates results.
This is especially sensitive in fine art and in communities that value mastery. Even when AI is used responsibly, it can raise questions like:
- What part of this image is the photographer’s work?
- What part is the software “inventing”?
- Would the image exist in this form without the tool?
5) Cultural and industry anxiety
There’s also a broader undercurrent: fear of being replaced, devalued, or drowned out by high-volume synthetic imagery. Even ethical, limited AI use can get swept into that anxiety—because people don’t always separate “AI to speed up masking” from “AI to fabricate entire scenes.”
A practical ethical framework (how to decide what’s appropriate)
When students ask, “Is it okay to use AI?” I think the most useful answer is: it depends on context and intent.
Here’s a straightforward framework you can apply:
Ask 1: What is the image claiming to be?
- Documentary / journalistic / natural history claim: authenticity is central. Extensive AI changes are usually inappropriate.
- Fine art / interpretive: more flexibility, but transparency still matters if the work is presented as a “photograph” rather than a composite.
- Illustrative / conceptual / marketing: AI use is often acceptable if it’s not misleading the viewer.
Ask 2: Did AI change the content or the meaning?
A helpful dividing line is: Did it materially change “what was there,” or just refine how it reads?
- Refinement: tonal correction, color balance, dust spot cleanup, subtle sharpening, minor distraction removal.
- Material change: adding/removing major objects, altering backgrounds, replacing skies in a way that changes mood or narrative, rebuilding structures, changing expressions/poses, inventing detail.
Ask 3: Who is the audience, and what are their expectations?
Collectors, editors, students, contest judges, and casual viewers often have different assumptions. Ethical practice means respecting the expectations of the space you’re in—or clearly re-framing the work so expectations align.
Ask 4: Would I feel comfortable describing my process out loud?
This is a surprisingly good test. If you would hesitate to explain what was changed, that’s usually a signal that disclosure is warranted—or that the edit conflicts with the norms of the context where you’re showing it.
My stance on transparency (and how I work)
This matters, and I want to be clear.
1) Be transparent when AI is used extensively
If AI-driven editing substantially changes the scene, structure, content, or meaning of the image—beyond normal tonal work and cleanup—viewers deserve to know. That’s not about shame; it’s about trust and context.
A simple disclosure line is usually enough.
2) I do not use extensive AI editing in my fine art photographs
For CFH Art, my fine art photographs are built through in-camera craft and traditional post-processing decisions—tone, contrast, color or black-and-white interpretation, and careful finishing—without using AI to invent or substantially reconstruct content.
That matters to me because my collector-facing work is rooted in authenticity: what I saw, how I framed it, and how I interpret it.
3) I will use AI tools for marketing materials
Where AI does make sense for my workflow is business support: mockups, promotional graphics, layout experiments, and other marketing assets. Those are communication tools, not documentary photographs. Used responsibly, they save time and help me show up consistently for students and collectors.
Suggested disclosure lines you can copy/paste
-
When meaningful AI edits are used:
“Note: This image includes AI-assisted editing to substantially modify elements of the original scene.” -
For marketing assets:
“Mockup created for marketing purposes; not a literal representation of scale or the exact display environment.” -
For an About/FAQ page:
“My fine art photographs are created through in-camera craft and traditional post-processing. I may use AI-assisted tools for marketing layouts and mockups, and I disclose AI use when it materially changes the content of an image.”
Bottom line for students
Used thoughtfully, this workflow can be:
- a learning accelerator,
- a time-saver for repetitive tasks, and
- a practical assistant for marketing production.
But your eye, your restraint, and your ethics still define your work. Tools evolve. Trust takes longer.
I’m in the wait and see mode as to how I can best use this connection between ChatGPT and Adobe.
