Can Artificial Intelligence (AI) finally protect artists from getting their designs ripped off?

Eric Zhang.
Eric Zhang.

Currently, artist copyright infringement is especially widespread in apparel – from corporations that mass-produce t-shirts, to ordinary people printing a design on a coffee mug. But could AI change the prevailing tide by stopping copyright infringement at the source? Yes, according to Eric Zhang, Vice President of Engineering at Scalable Press, the world’s first AI-powered print and fulfillment company. Here, he examines how the world’s hottest technology could turn the tide in the war for artists’ rights.

“News reports accusing fashion and retail giants of design theft have cropped up like clockwork over the past decade. In 2016, more than 40 artists and designers called out Zara for plagiarism, and the company’s woes have continued into 2018. Just this month, Harley Davidson was awarded $19.2 million for trademark infringement on the part of SunFrog, one of the world’s largest printed t-shirt distributors. While this is undeniably a victory in the effort against intellectual property infringement, independent artists don’t have the same resources as a huge company like Harley Davidson. In order to stop the problem at its source, the industry must turn to artificial intelligence.

The struggles of independent artists. Combating the “copycat culture” is anything but simple. Unlike other forms of creative media like film and music, fashion is often exempt from copyright laws because it’s a functional item. While artists can protect their designs, tackling intellectual property rights (i.e. trademarks, copyrights, design patents) is a tangled and lengthy process that can require significant financial resources. Even when artists decide to sue for copyright infringement – another drawn-out, expensive course of action – they rarely, if ever, emerge victorious over the daunting legal forces that large retailers have at their disposal. Items that come into question at fast fashion locations are quite often off the shelves before a lawsuit hits.

Plagiarism is just as rampant among small-scale peers, where imitations can appear on marketplace sites within months of a design’s release. And while artists no longer face the same legal fortitude as when they’re taking on a major brand, the time and money needed to reclaim a design can easily cripple its creator. Also, the Internet is a double-edged sword for indie artists. It provides platforms where their work can reach eyes that were once unavailable, while also presenting a space where most things are perceived as “fair game” and are easily taken without attribution. Thus far, social media has been the sharpest weapon in indie artists’ arsenal. Posts arraigning big brands can spread like wildfire, picking up press in a matter of hours and shaming them into reconciling with the parties they’ve wronged. However, when artists are calling out their peers, drawing press and public ire will diminish greatly with the loss of a “David-and-Goliath” angle. As powerful as the people’s voice can be, it shouldn’t be – and can’t be – the foundation in the fight for protecting artists’ rights.

The emergence of AI. Prior to Artificial Intelligence (AI), all artwork was individually hand screened by apparel manufacturers to detect copyright and trademark infringement. As you can imagine, this was an extremely tedious process. With the aid of AI, we’re now able to create statistical models for evaluating large quantities of artwork, so only a small fraction needs to come before human eyes. The vast majority are automatically labeled and rejected or approved. After AI pinpoints copyrighted or trademarked artwork, sellers attempting to print those designs will have to find a new design, or work with the original creators to license their artwork. This helps cultivate steady, collaborative avenues for indie designers that elevate their profile and deliver their due compensation.

We’ve already seen AI used in the battle against copyright infringement to great success with YouTube’s Content ID system. Employing audio and visual fingerprinting to detect copyrighted material uploaded onto YouTube, rights holders are able to have the offending material taken down quickly, or even cash in on the ad revenue generated by claiming ownership of the video. YouTube uses Content ID for 98% of its copyright management, resulting in less than 1% of disputed claims and over $2 billion in ad revenue paid out between 2014 and 2016. AI’s role in defending independent artists is still in its early stages, but it’s prepared to set a new global standard for the design industry. With apparel manufacturer endorsement, online marketplaces should follow suit, and eventually big brands will be held accountable. The message is crystal clear – technology must be adopted as creative content explodes, or you run the risk of dying out.”

Here’s how AI could revolutionize the apparel industry:
Apparel manufacturers will be able to use AI to help them identify the work of independent artists. This will stop people who are stealing artwork by reproducing a design, whether they’ve found it on purpose or by accident.

AI will be able to pinpoint the work of smaller artists, even if it’s not attributed. This will allow apparel companies to reach out to tomorrow’s design “superstars” for a collaboration – and sell even more merchandise!

AI will help the apparel industry set a global standard for other art-enabled industries, such as online marketplaces, by protecting artists’ rights and fostering new talent. Or, they’ll die out.

Scalable Press' combined facilities include over 100,000 square feet of production space.
Scalable Press’ facilities include over 200,000 square feet of production space.

About Scalable Press. With three locations in California, as well as facilities in Texas, Indiana and Pennsylvania, Scalable Press specializes in building “the next generation of printing and fulfillment technologies.” Its goal is to make custom merchandise fast, affordable and accessible at any scale. Its facilities include over 200,000 square feet of production space, 40 screen presses, 60 DTG (Direct-To-Garment) printers, and full-colour dye-sublimation printing. Its Engineering team, of whom Zhang is Vice President, is responsible for creating and iterating on APIs and production automation processes that power over a million printing impressions. The company is also focused on creating software that meets the flexibility and scalability needs of custom printing businesses.


Tony Curcio is the editor of Graphic Arts Magazine.