I used ChatGPT to respond to the following long prompt. A “conversation” ensued, part of which is copied below.
PROMPT: Camera clubs for amateur photographers have rules governing the photos that can be shown. These relate specifically to the authenticity of a photographer's images. Plagiarism is not allowed. Typical editing of digital images before the advent of AI was acceptable, but replacing the sky in an image with a different sky was acceptable, provided that both images were the photographer's own. Now, with generative AI, there is a need for more rules. Primary among these is the rule that an image, or part of an image, must not be created with a text prompt; after that, it becomes unclear to me. Many people accept that distracting parts of an image can be removed, provided that the area inserted is based on pixels from the surrounding areas. Traditional cloning and healing tools worked in this way, but with AI, it is unclear if that is the case. In Lightroom Classic, I have found that the remove tool without AI will sometimes make a mess of the removal, but with "Use generative AI" and "Detect Objects" selected, the result can be much better. The question is, did AI 'invent' the new area, or did it use pixels from the image? Super-sizing is another area of debate. Is the enlarged image a new image, or does it contain pixels derived from the original? In Photoshop, Generative Expand creates a new extension of the image that appears to be based on the original, but it is not the photographer's work. What guidance can you provide based on the documentation found online?
RESPONSE: A small but growing number of clubs and competitions are beginning to address the AI problem by separating competitions by editing philosophy rather than trying to police specific tools. This approach removes much of the ambiguity surrounding AI tools.
The model typically uses three clearly defined categories.
1. Straight Photography (or “Nature / Documentary”)
This is the strictest category, designed to preserve photographic authenticity.
Typical rule
The image must represent the scene as photographed. No elements may be added, removed, or relocated.
Allowed adjustments
exposure and contrast
colour balance
cropping
sharpening
noise reduction
minor dust spot removal
Not allowed
removing objects (people, branches, signs, etc.)
compositing
sky replacement
generative AI
significant cloning or healing
This model is widely used by organisations such as the Photographic Society of America and many nature competitions.
2. Altered Reality / Creative Photography
This category allows extensive editing and compositing, but still requires photographic origin.
Typical rule
All visual elements must originate from photographs taken by the entrant.
Allowed:
compositing multiple photographs
sky replacement (from the photographer’s own images)
object removal
creative colour grading
extensive cloning
Still not allowed:
text-prompt AI image generation
generative AI objects
stock images
images captured by someone else
The philosophy is:
creative manipulation is allowed, but the raw material must still be photography.
3. Open Digital Art / AI Art
Some organisations are now adding a third category.
Here almost anything is allowed:
generative AI
text prompts
stock elements
illustration
heavy digital painting
This category recognises that AI imagery is becoming a legitimate creative medium, but it is no longer treated as photography.
Why this system works
Instead of debating specific software tools, it focuses on creative intent.
Straight: Faithful representation.
Altered Reality: Creative photography.
Digital / AI Art: Image creation.
This structure prevents endless arguments about individual tools like:
generative remove
content-aware fill
AI upscaling
because the category's intent is clear.
An example rule set some clubs are adopting
Category A — Authentic Photography
Minimal editing only; the scene must remain unchanged.
Category B — Creative Photography
Extensive editing allowed, but all elements must be photographic captures by the entrant.
Category C — Digital / AI Art
Any digital techniques allowed.
Why this is becoming necessary
AI tools are improving extremely quickly. Within a few years, it will be very difficult to tell whether:
an object was cloned away
an object was generated
a sky was real or synthetic
By separating categories, clubs avoid the need to detect AI usage.
An additional rule that simplifies judging
Some competitions now include a declaration such as:
“Entrants must be able to supply the original capture file(s) for all photographic elements used in the image.”
This protects authenticity in the first two categories.
✔ In effect, the emerging international standard is:
photography categories → no AI-generated scene content
AI categories → anything allowed
This preserves photography while acknowledging that AI image creation is a different medium.