I read something recently that said people who use AI for their graphic design and marketing were never going to be your clients anyway.
It makes sense in some respects.
Most of the AI flyers you see across social media are being used to promote small businesses that would have previously used Microsoft Publisher to create their marketing. In other words, these aren’t necessarily people who were paying a professional designer in the first place. They’re either people who don’t have the budget for professional services or simply don’t place enough value on professional design to pay for it.
It’s not much different from cutting your own hair. Either you’re trying to save money, have some skill to do it yourself, or just don’t care that much about how your hair looks.
How is using AI design any different?
But then I had an actual client, one who I’ve worked with for years, send me AI-generated billboard artwork that he prompted himself and ask me what I thought.
It would never have been printable as he presented it.
It was the wrong size, about one-sixth the resolution it needed to be for billboard artwork, lacked bleeds and was cluttered as all get out with fonts that would never be legible by passing cars.
I have no doubt that this same client would have ended up paying someone at the billboard company to rework it into a printable version.
Instead, he paid me to do it.
In some ways, AI is blurring the lanes of professional services. It’s giving a false sense of empowerment to professionals who are REALLY good at their respective jobs and enticing them to step out of their roles into one they have neither been trained for nor have experience in managing.
In a lot of ways, using a tool like AI to create billboard artwork when you have no experience in print design is about as insane as thinking you can dig a footer for a house just because you bought an excavator.
Just because you have the tool doesn’t mean you have the knowledge to use it correctly for certain tasks.
Whether we’re talking about a house or a brand, both are going to be unstable if you try to build them without the appropriate skills.
So, here are seven reasons I don’t think AI is replacing professional graphic designers anytime soon.
1. AI Is Limited by the Engineer
AI can only be as useful as the person directing it.
Many people who call themselves “marketing specialists” don’t understand professional print preparation. They might be great at marketing. They may understand social media, advertising, sales or business strategy incredibly well.
But that doesn’t mean they understand graphic design and prepress production.
Someone without graphic design experience may not know to tell AI that a piece needs to be created at specific dimensions. They may not know the resolution required for the application. They may not know what bleeds or trim margins or creep are. They may not know when something needs to be built in CMYK instead of RGB.
They may not even know to ask.
And if you don’t know what you’re supposed to ask, you won’t be able to tell when the file you’ve been given is all wrong.
2. AI Can Muck Up Your Branding
I’ve seen AI take a perfectly good logo and completely screw with it.
Even when the actual logo is uploaded, I’ve watched AI distort the wording until it’s illegible, change branding colors, skew proportions or recombine elements in ways that were never authorized.
It matters a lot more when you’re talking about a business’s actual brand.
A logo isn’t just an image you slap onto something.
It represents the business.
There are reasons a professional designer establishes brand guidelines. You don’t simply let an AI engine reinterpret the logo every time you need a new piece of artwork.
And if you work for a larger firm, such as in real estate brokerage, your misuse of the corporate branding can have other repercussions professionally.
Consistency matters.
And so does knowing when you should NOT alter something.
3. AI Is Often Wrong
AI gets things wrong… a LOT.
It gets dimensions, facts, contact information and links wrong. It’s prone to making assumptions, inventing details and producing things that look perfectly plausible while being completely and utterly incorrect. Don’t believe me? Try asking it to cite references and data the next time it makes a claim that seems factual.
I’ve experienced this myself many, many times.
And this is where the engineer using AI still has to know enough to proof what it produces.
If you don’t know something’s wrong, you won’t fix it. You may not even realize there’s a problem until someone else has to fix it for you, or even worse, points it out publicly once it’s published.
AI doesn’t eliminate the need for quality control.
It makes quality control even more important.
4. Most People Don’t Understand Professional Files
Here’s another problem with the idea that anyone can simply use AI to replace a graphic designer: most people who don’t work in design don’t even know what files they need.
And that’s not a criticism. Why would they? I honestly couldn’t tell you what gauge wire an electrician needs for a particular application. I might be able to look it up if I had any understanding of electrical wiring at all, but it would be a lot faster, cheaper and safer to let a professional handle it.
If you’ve never worked with a printer, sign company, embroiderer, promotional products company or web developer, you probably don’t know the difference between a JPG, EPS and SVG.
You may not know what a vector file is, why it’s important or why the logo that looks perfectly fine on your computer screen might be completely useless for some of the things you want to do with it. You may not know why a printer, embroiderer or web developer is asking for a particular file, or when you need CMYK versus RGB.
The whole “AI can do it” argument falls apart for me when you consider this knowledge gap between the engineer and the tool.
I subscribe to professional software like Adobe and use its in-app AI tools myself on occasion. But I’ve also been working with these programs for more than two decades. I know what I’m looking at when I open a vector file. I know when paths are overly complicated, curves are sloppy, spacing is wrong… I know when something needs to be redone and how to fix it.
Someone without that experience may not know any of those things.
Having access to a sophisticated design tool doesn’t give you the knowledge to use it professionally.
5. One File Is Rarely Enough
When you design something with AI, you’re generally working with one generated option at a time and one resulting file or image.
That’s not how any professional designer approaches branding and marketing.
When I design a logo, for instance, I give my client a range of initial ideas. We look at those ideas, refine them and develop the direction that makes the most sense for their business.
And when I supply the final files to the client, I average a minimum of 90 total files.
I’m not kidding.
Those files include full-color, black and white versions, different layouts and orientations, and a range of formats and file types so the client basically has every version they may need of their logo for professional marketing and printing.
That’s part of what they’re paying for.
They aren’t just paying me to make an image that looks good on a screen. They’re paying me to think about all the possible ways that logo may need to be used throughout their business marketing.
Will it go on a website?
A billboard?
An embroidered hat?
A truck wrap?
A tiny favicon?
A professional designer thinks about those applications while developing the brand, not after the fact when a printer or vendor tells the client, “We can’t use this file.”
And there’s another part of this that I think gets overlooked.
I’m not simply taking an existing idea and asking AI to make it prettier.
I’m bringing ideas to the table that my client may never have thought of themselves.
I’m thinking about their competitors and their values. I’m thinking about how color, font and imagery play into human psychology. I’m thinking about how the logo needs to work now and what they may need from it five or ten years down the road.
That’s creative problem solving, not just file generation.
6. I Spy AI-Generated Design
As a graphic designer who has been doing this for decades, I can almost always tell when a flyer or other piece of marketing artwork was generated entirely by AI.
And sometimes I can even tell which AI engine created it.
There are distinctive hallmarks.
The problem isn’t necessarily AI itself. The problem is what happens when someone with no real design experience generates something with AI and assumes the first thing it spits out is finished artwork.
I recently counted 17 icons (the little line art graphics you see next to a phone number, email address or website) in an AI-generated flyer that was presented to me for my professional opinion.
I’m not kidding.
Seventeen.
That kind of clutter is highly typical of what I’ve seen ChatGPT produce when someone asks it to create a flyer and doesn’t understand basic principles of design.
It had terrible fonts. It wasn’t the right size for the publication. It had no bleeds.
And it used AI-generated photographs that were obviously AI-generated because the clouds, trees and houses in the images were far too uniform and unnatural.
The worst part was this flyer was promoting “luxury” properties while its design screamed anything but.
A professional designer knows that good design isn’t about figuring out how many things you can cram onto a page.
It’s about figuring out what actually needs to be there, establishing hierarchy, knowing what someone should see first, second and third, and frankly knowing when to just leave something out.
It’s also about typography, spacing, balance, contrast, readability and making sure the whole thing can be understood in the few seconds someone actually gives it.
Good design should always be distinctive.
Every time I open up a social platform now, I see AI-generated flyers that, at a quick glance, look like they could all be advertising the same business. One might be a car detailer, another advertising a churchwide picnic, while two more nearly identical flyers might be promoting a new restaurant in town alongside a local exterminator. They’re completely different businesses with different audiences, yet somehow, they all look like they came from the same design template. And this right here is the irony of using AI because you want your business to stand out, yet you end up looking like everyone else.
Could someone with design experience use AI to create something much better? Absolutely. The better the person prompting the AI understands design, the better they can direct it, recognize what’s wrong and know what needs to be changed.
But that person isn’t getting professional results because AI magically made them a designer. They’re getting better results because they already understand design.
AI can produce something that looks finished. A designer knows whether it actually is finished.
7. Ethical Questions About Creative Material
There’s a very real and ongoing debate about the ethics of AI engines “stealing” creative material and scavenging it for parts.
I’m intentionally using the word debate here because the legal and ethical questions surrounding how AI companies obtain and use creative work are still being worked out.
There are ongoing lawsuits, legislation and disagreements over copyright, consent, attribution and whether and how copyrighted creative work can be used to train AI systems.
But even if you aren’t a graphic designer, these questions should matter to you as a business owner.
When you engage with any entity that is potentially using creative material in ways that may ultimately prove to be illegal or at least ethically questionable, you have to consider how that could affect your own reputation.
It’s a little like a pawn shop buying stolen goods.
The pawn shop didn’t necessarily steal the goods itself. But if it turns out the goods were stolen, the fact that the pawn shop participated in the transaction may come with consequences.
And if you think you’re protected just because Google created it, think again. Google’s terms specifically address how it handles certain intellectual property claims and place responsibilities on the user.
Responsible businesses use discretion and wisdom when sourcing their suppliers, especially when their own customers may have strong feelings about these issues.
If your company prides itself on supporting artists, photographers, writers or other creative professionals, what happens when one of your customers discovers that the imagery you’ve used was generated from a system whose training practices they strongly oppose?
You may not consider it a problem, but your customer might find it hypocritical and unethical.
And in a world where consumers increasingly care about how the businesses they support operate, that’s a reputational issue worth considering.
I’m Not Anti-AI. I’m Pro-Expertise.
I’m not anti-AI.
I believe it can be a useful tool in many respects and use it in different ways both professionally and personally.
In fact, I’m using AI to help me write this article, which I realize is pretty darn ironic.
I majored in Journalism. I researched this. I wrote the overwhelming majority of it. I edited it. I’m using AI as an organizational and revisionary tool, not a substitute for thinking. These are my words, my thoughts, my research and my real-life examples.
But there’s a difference between using AI as a tool and believing the tool has eliminated the need for expertise.
They aren’t just paying for access to design software or AI. They’re paying for the years of experience behind the person using those tools.
That experience means knowing what questions to ask, what needs to be fixed, what will actually work in the real world and what looks good on a screen but is completely wrong for the application.
The point is that good design has never been about having the right tools.
It’s about knowing what to do with them.
Give a professional designer an AI tool and you’ve given her another tool in her toolbox.
Give someone without design experience the same tool and you haven’t given them twenty years of experience, a working knowledge of prepress, an understanding of typography, branding, visual hierarchy or the ability to recognize when something looks good but is completely wrong.
You’ve just given them an excavator.
And an excavator alone can’t dig a proper footer.
