There is something that constantly happens in my family. My sisters—especially the second oldest—love sending old family pictures into the family chat. Every once in a while, I’ll look at one of those faded photographs and think, “Let’s throw this through ChatGPT and see what it can do.”
People tend to think restoring old photographs with AI requires complicated prompts and technical knowledge. In reality, some of the most effective commands are also the simplest.
The first command I almost always use is:
“Color correct this image.”
Think of it as the magic button.
You’re asking the AI to look at the photograph and balance the colors so skin tones, clothing, walls, furniture, and everything else look more natural. It often fixes the yellowing, blue tint, or faded appearance that develops as photographs age.
The next command is usually:
“Make it a little brighter.”
Notice the words a little.
You are trying to restore the image, not make it look like it was taken yesterday. Too much brightness or too much contrast can wash out details and remove the character that made the original photograph special.
From there, you can experiment with commands like:
- Add a little contrast.
- Recover shadow detail.
- Sharpen slightly.
- Remove dust and scratches.
- Preserve the original faces.
- Keep the image looking natural.
The AI usually gives you a solid starting point. Then you decide whether it needs another pass.
Photography Was Always More Than Pressing the Shutter
One thing people sometimes forget is that photography was never simply about taking the picture.
There was an entire craft behind producing the finished print.
Back in the days of film photography, photographers weren’t using just one type of paper. Proof prints were often made on less expensive paper because their purpose was simply to let customers choose which photographs they wanted to purchase.
Once the customer placed the order, the final print could be produced using better photographic paper and the finish they preferred.
Some customers liked glossy prints because they made colors appear more vibrant.
Others preferred matte finishes because they reduced glare, hid fingerprints better, and were easier to display under normal room lighting without reflections bouncing off the glass.
Photography involved choices every step of the way.
Knowing the Chemistry
Something else that has largely been forgotten is that photography was chemistry.
A lot of the quality of a finished print came down to knowing the process.
Darkroom technicians understood that proper fixing, washing, and finishing could dramatically improve how well a photograph survived over time. Many people never saw those extra steps because they simply bought the finished picture.
There was a tremendous amount of trade knowledge behind producing photographs that would last for decades.
That’s something most people today never experienced.
Every Medium Ages Differently
Then you have all the different photographic formats.
Traditional prints age one way.
Polaroids age another.
Negatives have their own challenges.
Slides can be particularly interesting because you’re restoring film rather than paper. Colors shift differently over time, dust becomes more noticeable, scratches appear more easily, and converting a slide into a modern digital photograph often requires additional correction before you even begin restoration.
Each format brings its own personality—and its own problems.
That’s why there isn’t one perfect AI prompt that fixes everything.
AI Doesn’t Replace Judgment
One of the biggest misconceptions about AI photo restoration is that you press a button and you’re done.
That’s rarely how it works.
AI gives you a starting point.
Sometimes the first result is excellent.
Other times it over-brightens the image, changes facial details, or makes colors look unnatural.
The important part is looking at the result critically.
Did it preserve the people?
Did it preserve the expressions?
Did it keep the photograph feeling like the original memory?
If the answer is yes, you’ve probably done a good restoration.
If not, give it another instruction.
The New Darkroom
In many ways, AI has become today’s darkroom.
Photographers once adjusted chemistry, selected photographic paper, chose matte or glossy finishes, and carefully processed prints to produce the best possible image.
Today we’re adjusting prompts instead.
“Color correct.”
“Brighten slightly.”
“Restore detail.”
“Remove scratches.”
The tools have changed.
The goal hasn’t.
Whether you’re working with a faded family portrait, a decades-old Polaroid, a box of slides from the attic, or a stack of aging prints, you’re doing the same thing photographers have always done—trying to preserve memories so they can be enjoyed by the next generation.
Sometimes the simplest prompt is still the best place to start.