Tag: AI

  • Looking To The Future: The unclear legal landscape of AI images

    Picture this: A new customer enters your printshop and asks you to print a thousand shirts using an image they have provided. 

    That image clearly depicts a character you recognize: Darth Vader. 

    You can tell that this image was created using artificial intelligence (AI). You also know the customer likely does not own the intellectual property or hold the copyright to the Darth Vader character. 

    This job request creates a moral, ethical, and even legal dilemma. 

    Do you accept the job to print a thousand shirts? Or do you inform the customer that since this is clearly a copyrighted image and they don’t own the rights, you can’t take on the job? Do you have a legal responsibility to turn the job down? 

    This scenario is happening more often in print shops nationwide and worldwide now that AI is used regularly to create images. Printers can get ahead of the dilemma by considering the hard questions first. 

    The Darth Vader Dilemma

    As a business owner, you might struggle inwardly with your own business ethics, but the real question is your legal responsibility. Since the image was created with AI, does that mean your customer owns the rights to it? 

    The United States Copyright Office currently holds that copyright requires some aspect of human authorship. 

    In other words, dictating prompts into an AI platform to create an image is not enough to establish human authorship and qualify for copyright protection. 

    But that doesn’t mean that anything made with AI can’t be copyrighted. The Copyright Office also says human-created AI images can potentially receive copyright protection. That distinction essentially depends on whether AI is a tool or the creator of a human author’s expressive choices.

    The Darth Vader image raises another trademark issue. Darth Vader as a name is a legally trademarked entity owned by Lucasfilm, a Disney subsidiary, and the character itself is protected by copyright.

    Deciphering the Dilemma

    Marshall Atkinson operates two businesses from his home base in Mesa, Arizona. One is Atkinson Consulting, and the other is Shirt Lab Tribe. He also runs a side project called the Midjourney Experience Academy, where he teaches people how to use Midjourney, an AI platform. Atkinson says he uses his preferred AI platform daily, and he keeps up with the ever-changing technology, trends, and even legal disputes.

    Midjourney has already had one major copyright-infringement ordeal. In June 2025, Disney and Universal sued the AI platform in federal court in California for copyright infringement. They alleged that Midjourney copied their images to train its system. 

    The plaintiffs also alleged that Midjourney produced and distributed images that incorporate their protected characters. While this might seem like an open-and-shut case in Disney’s favor, Atkinson has a different point of view.

    Atkinson contends that the legal lines are blurred when it comes to using specific prompts in Midjourney. Certain names and phrases are blocked as prompts, such as “Mickey Mouse” and “Donald Trump,” but others, like “Darth Vader,” are not. 

    “If it’s such a big deal, why not block everything mentioned in the lawsuit?” he asks. “Let’s take AI out of it, because AI only makes art generation easier. Why do they allow all these people to dress up like Star Wars characters and go to the convention or do their own fan art?” 

    In this instance, Atkinson argues most people aren’t using AI tools like Midjourney for commercial works. 

    Lisa Higginbotham, president and owner of FiveStar Awards & Engraving, says that businesses have a responsibility when it comes to accepting AI-created images for print jobs. She says businesses should do their best to avoid printing other people’s intellectual property. 

    “You cannot print a Darth Vader image because Disney owns it,” she argues. “People do it because they think they’re too small and that they won’t get caught. But it only takes one person turning you in or you posting it on social media, then it gets to Disney, and you’re out of business. I’ve been doing this for 26 years, and I don’t want Disney knocking on my door saying. ‘Shut it down.’”

    While Atkinson believes users should have more freedom with AI creation, he agrees that print businesses need to be wary of accepting jobs with AI-created artwork. 

    “For a business, you have to have your own set of rules,” he said. “This is your ethics and what you do. I wouldn’t use other people’s IP because you could be open to a lawsuit.”

    A Different Perspective: From User to Developer

    Meanwhile, on the other side of the world, Chris Minn operates his company, Digital Ink, in Muscat, Oman. 

    Digital Ink is a hybrid marketing collective focused on consultancy and AI development for print-software vendors and industry leaders around the world. Minn and his business partner are set to launch their own AI platform called GEORank, built to help print service providers sell their products in the AI world. 

    “GEO” stands for generative engine optimization, and the platform will use it to rank printers in search results. Minn says more people use AI over traditional search engines like Google now.

    “You may have been a really good print shop for 30 years as a family-run business through recommendation, word of mouth, and Google results,” he states. “But if 50% of Google search is now gone, that’s 50% of people that won’t find you.” 

    Businesses that don’t adopt a GEO strategy now could potentially be outmatched by a competitor simply because they’re more organized with their content, he adds. 

    Minn previously worked for a web-to-print B2C gifting company and says he understands how AI fits well into the personalization segment of the print industry.

    “The whole point of personalization is being able to give someone something truly one-off, instantly and affordably, and that’s just not something a human design team can do at scale,” he suggests. “AI fills exactly that gap.”

    But that efficiency and convenience can come at a cost. Without checks and balances, Minn says business owners could find themselves in legal hot water for inappropriate or unlicensed images. 

    As someone following the AI-legality debate, Minn has watched several lawsuits closely and stressed that two separate conversations are happening around AI-generated artwork: 

    1. Ownership: whether you can copyright something you created with AI. 

    2. Copyright: whether using an AI tool exposes you to someone else’s copyright claim. 

    Minn says he looks to the U.S. Copyright Office and ongoing U.S. lawsuits for guidance in his work. He says he believes the U.S. is leading the global legal conversation on AI. 

    Navigating the Unclear AI Pathway

    As the courts continue to debate these conversations, it’s still crucial for business owners to build contingency plans for themselves and their customers. 

    Keep it human: Atkinson suggests uploading a photo or hand-drawn sketch onto the platform rather than simply typing prompts. Overall, he believes the use of AI is a personal decision.

    Maintain a record: Minn concurs that adding a human touch to AI image creation could be a safeguard against copyright-infringement lawsuits. He says his own pathway to avoid AI legal jeopardy comes back to a handful of habits, such as keeping a record of the human work that went into the design, knowing and understanding the AI vendor’s terms, and keeping the prompt history and source files used during image creation.

    Establish boundaries: To avoid headaches with unlicensed images or intellectual property, Minn suggests adding a legal message to automated upload systems and requiring submitters to acknowledge that the image is their original artwork. An attorney should also be consulted for this step, since language and laws differ state to state. He said business owners should also keep up-to-date on the U.S. Copyright Office’s reports. 

    Building New Standards 

    Minn contends that from a strictly legal perspective, there needs to be a recognized and enforced industry standard for AI-image usage so that filters can be universally put in place to avoid criminality for copyright infringement and trademarks. 

    “My father always said to me, ‘The internet is one of the best and worst things ever,’” Minn said. “It gives you so much information you can achieve anything, but also criminals will manipulate it as well. 

    “Until we get some kind of international industry legal standard [for AI], we run the risk of those filters not being on all large language models,” Minn concludes.

  • Three Favorite AI  Tools in Photoshop

    Looking for a new read that may give you some tips for business and life? A few APA members share books that inspired. 

    By Shon Roti

    AI

     impacts nearly everything we consume and create, and there are no signs of it slowing down. Graphic design and design software are increasingly shaped by AI-driven tools. 

    My own resistance to using AI in daily design work has gradually faded as I have become more aware of its practical benefits and time-saving capabilities, while also coming to terms with my initial concerns about what it might mean for my relevance as a marketing professional.

    AI in Photoshop 

    Currently, there are more than 25 AI-powered features and tools in Adobe Photoshop, plus at least 20 machine-learning-assisted functions. Their purposes include adding or extending scenery; removing, adding, or selecting people or objects; retouching skin blemishes; restoring photos; and adjusting lighting, to name a few.

    Here’s a look at a few AI features I often use to save time. 

    Figures A and B: Two photos taken on different days with very different lighting situations: midafternoon and cloudy for the red Nova wagon and sunny late afternoon for my green FJ Cruiser. 

    Figure A

    Figure C: A composite created using three different AI tools and surprisingly few steps to complete this convincing effect.

    Figure B
    Figure C

    Select Subject and Generative Fill

    Access the Select Subject option in the Contextual Task Bar by right-clicking in the empty workspace (not the image) and selecting Show Contextual Task Bar (figure D).

    Figure D

    To select the most prominent object in the photo, use Select Subject. Depending on the computer’s processing power, this could take as little as five seconds and often does a better job than I could have done manually with the Polygonal Lasso Tool.

    Next, in the same Contextual Task Bar, the Generative Fill option appears. I selected it. At that point, I could have chosen to fill the space with whatever I wanted: a fruit basket, twelve monkeys, or another random vehicle. But in this case, I entered the prompt, “Remove the car.” 

    Photoshop completed the task in seconds, leaving an empty space in the parking lot (figure E). It was not perfect; there was a bit of residual shadow left behind from the Nova, but it was still pretty impressive.

    Figure E

    I then right-clicked the newly created and merged empty-parking-lot layer and merged it down into the layer below (figure F).

    Figure F

    Next, I opened the photo of my FJ Cruiser, used the same AI selection option, copied it (Ctrl+C), and pasted it (Ctrl+V) into the parking lot file, above the layer containing the empty parking lot (figure G). The FJ and the background looked about as natural together as a fish on a bicycle. 

    Figure G

    Harmonize Layer

    You’ll locate this feature with the search bar in the upper-right corner of the workspace. Type “harmonize” into the search field, and the Harmonize Layer option should appear near the top of the results.

    After confirming I had selected the FJ Cruiser layer, I chose the Harmonize Layer option (figure H). This process took a bit longer, but it adjusted the lighting to match the cloudy conditions of the image below. However, it also created a new shadow under the FJ and partially removed the shadow from the hood. It also gave me two other renderings to choose from. I chose the first option. Then I merged that layer into the one beneath (the parking lot). I copied the new image, pasted it into a new document, and saved it; the illusion was complete.

    Figure H

    Of course, I would be remiss if I didn’t also show how these new AI features can be applied in a real-world situation, like the T-shirt design in figure I.

    As with most emerging technologies, some AI design tools will become more relevant than others, depending on the type of work or the project at hand.

    Full disclosure, AI was used to assist with definitions of these AI features, improve spelling and grammar, and generally polish my sentence structure. The photo of the lady in the white T-shirt was found on Pixabay.com and is an AI-generated image. However, the design and the article itself still come from human skills, honed over time.

    Figure I

    Like everything in life, there are few guarantees, but one of them is change. I am not ready for retirement yet, so I am adapting and learning at least until the robots take over.