Key takeout
• Generation AI is increasingly used in the promotional product industry, with 24% of suppliers and 19% of distributors adopting it for artwork or video generation.
• AI art generators can dramatically increase output and quickly process complex styles, but struggle with text accuracy, realistic embroidery designs, and subtle and emotional elements of famous logo creation.
• Experts cite potential unemployment, high energy consumption and copyright risks as important concerns, urging the call to AI transparency and thoughtful integration.
The sparkling geometric puzzle pieces fit perfectly behind the company logo. Moss-Green Floridian Fliage is the background of a dive letter promoting the family festival. The grinning raccoon palms a foamy beer next to a witty catchphrase.
Few people speculate that the graphics were created with the help of generator AI.
Marshall Atkinson, a veteran of the decorated apparel industry, partially created the creation using Midjourney, an AI model specializing in images. He says that AI-generated artwork can reduce his design time and replace up to 80% of his creative process.
“Why do I spend three hours doing something when I can spend 20 minutes?” he asks.
Atkinson used the generated AI to create the foundations for these designs, tweaked them, and added colors, text and logos to create the final product.
Although Atkinson says he still does it manually to add elements like text, color, and company-specific design, he is part of an increase in designers using AI artwork for creating logos, brainstorming, or other means of streamlining the creative process.
According to a 2025 industry report, 24% of suppliers and 19% of distributors use AI for artwork or video generation. This is the third most common reason why we use technology for our suppliers and distributors.
Generic AI can be a powerful tool that helps you save time and money. Experts say, especially for promotional companies with low staff or tight budgets. However, some industry experts are more cautious about the technology by citing copyright concerns and changing consumer expectations.
“To be honest, I don’t think that using AI is what (industry experts) can use AI,” says veteran embroidery digitizer and designer Erich Campbell. “There are many uses that don’t need to hinder one creative thing we are doing.
Possibility of AI promotion
One of the greatest benefits of incorporating AI into art creation is the speed at which designers can iterate through new ideas and create mockup designs, proponents say. For example, last week, freelance artist Michelle Moxley used AI to create 60 T-shirt graphics in two days. This is an impressive productivity boost compared to traditional methods. Relying primarily on clip art, Moxley says you can create three to five graphics a day. Using hand-drawn sketches reduces that number to 1-3.
“AI offers more diverse options and allows for more frequent updates, which increases overall value,” she adds.
But Moxley points out that treats AI like a “custom clip art machine” and creates graphics under different generators, cuts them out and puts them together.
“It’s a bit of a collage effect,” she says. “It’s not just me going and generating a design and then taking it. I take the piece.”
Moxley created these T-shirts with the help of AI. This has made the design rooted in real places.
In addition to increasing the amount of artwork created, a well-trained AI model can take on complex tasks at the speed of lightning, says Moxley. She trained AI models to replicate seed art, a technique for creating complex mosaic style designs using seeds.
“You can make seed art, but can you make 25?” Moxley says. “It would take me two years. This would take 20 minutes if I trained it.”
Moxley trained AI models by supplying specific data and creating artwork that mimics seed art techniques.
AI tools are also used to support print-on-demand production. NJ-based supplier Unionware (ASI/73775) recently collaborated with New Jersey Institute of Technology to develop an AI-powered pattern generator for its tote bags. Customers can choose from hundreds of pre-generated options, edit existing AI prompts, or create their own designs. According to Unionware president Mitch Kern, you can also upload existing patterns (probably printed on a T-shirt).
“First of all, what we’re trying to do is choose hundreds of designs for our clients. Then we can easily edit those designs so that we can fine-tune them into other colors and other patterns, and make it easy to do the designs they like,” says Cahn. “The next step is to help people guide them to educate them on how to come up with the best prompts.”
Limitations of art generated by AI
According to the Interactive Design Foundation, AI models such as Midjourney and Openai’s Dall-E have become increasingly sophisticated over the past few years, as neural network algorithms can continue to learn from large datasets of images and generate more realistic artwork.
However, AI still has limitations when it comes to spelling and replicating fonts, says the designer. Moxley’s podcast Fahrenheit AI explores AI applications for creative professionals, but it has won the name because only AI image generators created by former Google Brain Researchers can spell “Fahrenheit” correctly. Atkinson adds that he doesn’t rely on AI for text because he is sticking to fonts.
In the embroidery department, the artwork generated by AI for now is not a major player, according to Campbell. Some companies use Splashy Language to describe AI-powered digitization, but he says it is a buzzword that promotes the same automatic digit process that has been around for many years.
While AI-generated images to replicate the embroidered look may look promising, he adds, it is often made up of incredibly small stitches that are unrealistic to render in threads.
“We still don’t have an automatic way to photograph artworks that you give us and spit out quality embroidery,” Campbell says.
Despite its advances, AI is not a realistic alternative to logo design, especially for multi-million dollar companies with a well-known image, says Vladimir Gendelman, CEO of Company Folders (ASI/287795).
“When designing a logo, you need to consider the emotional aspects, the human touchpoint,” he adds. “AI doesn’t do that.”
The ethics of AI-generated artwork
In the August 2025 issue of Vogue, a smiling blonde model with white teeth, light eyes and smooth skin was the face of speculative ads. She didn’t look too perfect – because she wasn’t. After the brand confirmed that the models were AI-generated, readers killed criticism in fashion magazines. In the complaint? Using AI-generated images promotes unattainable beauty standards and takes jobs away from human photographers and models.
Designers and artists are already feeling the pinch of declining employment opportunities, as AI-generated artwork is very easy to produce and less than hiring artists.
“All our friends, people who are digital artists, who live on the committee, are suffering,” he says. “That’s an honest truth.”
He adds that there will always be opportunities for highly skilled artisans, but it may not be as unfeasible as before to vectorize the cleanup of a standing logo or basic image.
Another frequent criticism of AI artwork is its high environmental impact. Considering all the energy needed to run a computer, approximately 2,282 Joules of graphics processing units are required to generate standard quality images with the leading open source image generators. According to MIT Technology Review, a slight improvement in image quality and a slight improvement in image quality will increase an estimated 4,402 joules. Consider that users can quickly and easily create new images using AI. Easy to see the incredible energy costs of technology. By 2028, experts estimated that more than half of the electricity sent to the data center will be used for AI. AI uses about 22% of all US households each year, according to the same MIT article.
Another major mainstream criticism of AI-generated artwork is copyright infringement, with many authors, artists and companies claiming that large-scale linguistic models create copyrighted images based on trained materials. In recent years, dozens of copyright lawsuits have emerged from companies such as The New York Times, Universal Music Group and individuals such as authors Sarah Silverman and Ta Nehisi-Coates.
“People were torn through Mickey Mouse long before AI. When the photography began, all the portrait artists suddenly said, ‘Art is dead.’ Is art dead since photography was invented? – Marshall Atkinson
That’s part of why Moxley is transparent about when using AI. She works with other clients and other clients who have the potential to integrate technology and are resistant to it.
“Do you think the AI is stealing? Yeah, I do,” she says. “I don’t know how I feel about it because in a way there should be regulations, but I don’t know anything else.”
And Atkinson, who runs newsletters and websites that guide creative experts through AI, says he uses technology as a tool, not as a way to copy other people’s work. AI models train data, learn the colours of blue, how horses look, and the complexity of the sunset, not unlike students who recreate artwork in museums by studying their composition and structure.
“People were torn Mickey Mouse long before AI,” he says. “When the photography began, all the portrait artists suddenly said, ‘Art is dead.’ Is art dead since photography was invented?
