How AI Product Photography Workflow Is Rewriting Commercial Visuals

If you work with brands, you are already feeling the shift.

Suddenly everyone wants “AI visuals,” faster turnarounds, smaller crews and more versions. But very few people can explain a real AI product photography workflow that a serious brand would actually trust.

That is exactly what we explored in my AI Insights San Francisco conversation with Claire Xue, a creative leader who lives at the intersection of product photography, generative AI and brand work for companies like Sephora and top agencies.

“The commercial industry is going to be heavily impacted by this technology because brands are always looking to reduce the costs and the production lead time.” — Claire Xue

In this article, I’ll walk through Claire’s workflow, the business impact behind it and the ethical guardrails she’s putting in place so creatives do not get left behind by their own tools.

 

Why AI Product Photography Workflow Matters Right Now

Claire’s perspective is grounded in the reality of commercial sets, not theory.

She started in traditional product photography and set design, then moved into AI as she watched brand expectations change in real time.

“Traditionally you have to buy all these different props, go to different locations, rent studios, hire models. With AI-enhanced workflows you can still produce high quality content, but with a much shorter period of time.” — Claire Xue

When you zoom out, the value of a strong AI product photography workflow is pretty simple:

• Cut prop, location and studio costs

• Reduce total production time from weeks to days

• Make it easier to run A/B tests and personalized creatives

• Keep human craft and brand taste in the loop

The key is not to replace photographers or directors. It is to use generative AI where it makes the most sense in the pipeline.

Claire’s “Million Dollar” AI Product Photography Workflow

One of my favorite parts of the episode is when Claire walks step-by-step through the workflow she teaches her students and brand teams.

She calls it a hybrid AI product photography workflow because humans and AI share the work.

Step 1 – Generate a background and product “placeholder”

Instead of starting in camera, Claire starts in the model.

She prompts for an editorial-style scene with a fake placeholder product:

“I will use a prompt like: an editorial photoshoot of a glass of whiskey, a bottle of Hennessy sitting on a marble countertop. I describe the scene, the props, the environment, the color palette and the light.” — Claire Xue

The goal here is not perfection. The goal is composition and lighting direction.

Step 2 – Match the lighting in a real photoshoot

Once she has a background image she likes, Claire studies it like a traditional photographer:

• Where is the light source

• What is the angle and intensity

• How shadows are falling on the fake product

“Based on that, I will mimic the lights. I see where the light source is coming from, and I will take my product and actually shoot it with the same lights, settings and perspective as my generated image.” — Claire Xue

This is the first big unlock. Instead of guessing the set, she reverse-engineers it from the AI image.

Step 3 – Enhance and harmonize in post

After the physical shoot, everything comes together in post-production.

Claire enhances the AI background, then uses Photoshop (and now its generative tools) to blend the real product into the AI scene:

“The last step I always end in Photoshop, basically blending the actual photograph of the product into the AI generated scenery.” — Claire Xue

New features like harmonization make the process even faster:

“Recently they released the harmonization feature. You can just pop in the product photo you shot and it will harmonize the lights for you.” — Claire Xue

The result feels like a full-scale editorial shoot, but produced with a fraction of the traditional budget and lead time.

No wonder I called it a “million dollar workflow” in the conversation.

“There’s like a million dollar workflow right there.” — Roan Weigert

Cost and Time: Where AI Product Photography Really Wins

If you have ever produced a commercial shoot, you know exactly where the money goes.

  • Props and custom sets
  • Location fees
  • Models, wardrobe, hair and makeup
  • Producers and crew
  • Studio rental and retouching

When I asked Claire where AI makes the biggest measurable difference, she did not hesitate:

“I would say two things: cost and time. Traditionally you have to buy all these props, rent studios, go on location, hire models and full crews. With more AI driven or hybrid workflows you can cut those costs and shorten production dramatically.” — Claire Xue

For many campaigns, brands do not want to get rid of real shoots. They want options:

• Hero campaign shot in real locations with models

• Variations, social cutdowns and experiments created via AI workflows

That mix is where an AI product photography workflow shines.

Personalization And A/B Testing With AI Visuals

One area that gets Claire particularly excited is personalized ad creatives.

Instead of one hero image in one setting, AI allows brands to adapt visuals to different audiences and contexts.

“You can easily change the colors, change the background, personalize content for an ad or a post. Different people react differently, and now A/B testing with AI is much easier.” — Claire Xue

Imagine:

• Same product, but background adapts to local culture or season

• Matching color palettes to customer segments

• Testing 10 versions of a composition in days, not months

The infrastructure is still being built, but Claire is clear about the direction:

“That is definitely where we are heading. You and I could be sold the same product, but with different content that’s catered to our preferred way of consuming media.” — Claire Xue

An AI product photography workflow is the foundation you need before personalization at scale becomes standard.

How Big Brands And Agencies Are Experimenting

This is not just a creator trend on social media. Claire is seeing the shift from inside major companies.

“Early 2024 is when I realized there were a lot of sign-ups from major brands, top agencies, even auction houses wanting to learn how to integrate generative AI images into their product workflows.” — Claire Xue

Behind the scenes, teams are:

• Testing internal pipelines with tools like ComfyUI

• Building custom workflows for different product categories

• Running controlled campaigns that mix traditional and AI-driven assets

And it is not just static imagery. We talked about:

• Mango experimenting with AI-generated models for fashion

• Coca-Cola running fully AI generated ad campaigns

• AI film competitions like Chroma Awards aiming to showcase new visual languages

“You see brands like Mango launching collections with AI generated models, and Coca-Cola doing AI ads. That was a pivotal moment. From there, more brands started asking how they can use AI for campaigns.” — Claire Xue

If you are a creative director or marketer, this is the moment to learn the tools before your clients ask.

Ethics, IP And Responsible Prompting

With all this power comes a real risk of crossing ethical and legal lines, especially around training data and style copying.

Instead of trying to solve model training debates, Claire focuses on what users can control.

“I’ve always tried to shift the narrative from how the models were developed to how we as users and brands adopt these tools responsibly.” — Claire Xue

Her guidelines are simple but strong:

• Do not put artist names in your prompts if you do not have consent

• Do not use reference images you do not own or have rights to

• Keep humans in the loop for taste and accountability

• Attribute human creators involved in the workflow

“Personally I always suggest people and brands not name names in your text prompt, and not use reference images you don’t have rights or consent to.” — Claire Xue

She also emphasizes traceability:

“Always keep the human in the loop and build a system so you can trace back who contributed what. We still need humans to utilize these tools.” — Claire Xue

The result is an AI product photography workflow that is not just efficient, but defensible.

How Creatives Can Enter The AI Product Photography Space

If you are a photographer, videographer or designer, it is easy to feel threatened by AI tools. Claire feels the opposite.

“I realized if we don’t keep self-educating ourselves as a creative community, we’re at risk of letting tech companies dictate our creative future.” — Claire Xue

Her advice for creatives starting now:

  1. Start in your own niche

    “Build a portfolio in your existing domain expertise. That will guarantee you have the highest quality results when you’re using AI.” — Claire Xue

  2. Experiment towards a repeatable workflow

    Different use cases (product shots, virtual try-ons, lifestyle scenes) need different pipelines. Try multiple tools and find the flow that fits your clients.

  3. Share your learnings with your community

    Claire runs bootcamps and newsletters not just to teach, but to stay sharp herself.

    “I started a newsletter and a camp because I wanted to keep myself updated and at the same time keep my audience in the loop.” — Claire Xue

  4. Treat AI like a co-pilot, not a replacement

    You still decide the concept, taste and final call. The model just helps you get there faster.

San Francisco, Community And Being The Bridge

Even though San Francisco is dominated by LLM and B2B infrastructure companies, Claire sees a clear role for creative leaders in the city.

“I do feel like as a creative person, I have a responsibility to learn what is happening in the Bay Area and be the bridge between the tech and creative community.” — Claire Xue

By showing up at hackathons, AI events and creative meetups, she is:

• Translating technical advances into creative language

• Bringing real use cases back to brand marketers and artists

• Making sure creative voices are heard in how tools are built

That is also the mission of AI Insights San Francisco: to put founders, builders and creatives in the same room and let the real stories drive the conversation.

Final Thoughts: Build Your Own AI Product Photography Workflow

If there is one takeaway from this episode, it is this:

You do not need a giant lab or custom models to start. You need a clear, ethical, repeatable AI product photography workflow that fits your niche.

Start small:

• One product

• One scene

• One hybrid workflow where AI sets the stage and your craft finishes the shot

Then iterate.

As Claire put it, the technology is moving with or without us:

“This is the time to be in San Francisco. There are so many events and hackathons. Everybody is so brilliant and builder-minded. You constantly get inspired and motivated.” — Claire Xue

If you are a brand, creative director or independent creator, now is the moment to experiment, document your process and turn AI from a buzzword into a real competitive advantage.