Artificial Intelligence (AI) is everywhere, and it’s safe to say that most everyone is using it.
But should they be?
I’ve been asked a few times in the past year to take a logo that someone has generated from an AI platform, and to refine it – whether that means making the words look better, changing out the colors, or balancing elements that AI just couldn’t get right.
But here’s the thing: I have big-time ethical issues with using AI artwork. And this kind of thing has become enough of an occurrence in my business that I have to start talking about it more.
So what’s the big deal? AI can’t create something original or unique (not to mention the environmental and societal impacts of using AI tools). To the point that the United States Copyright Office and the U.S. Court of Appeals agree that AI-artwork cannot be copyrighted, because a human wasn’t the one creating it in the first instance. [Source]
What does that mean for businesses and branding?
The question I get most often when I bring this up is, “why should I care about copyright?” or the statement, “we don’t need to copyright our logo.”
This is a foundational misunderstanding, and it can cause problems in the long run for you and your business or organization.
Copyright is inherent for works for art – the person who created it owns it unless they transfer the rights. If someone were to use their artwork without permission, the creator (or rights owner) would have legal grounds to go and tell them to stop using it, and could take that as far as they needed or wanted to.
But when the “artwork” you create using AI isn’t copyrightable, you have no grounds. Which means you have no (legal) control over how your brand is used or by whom.
While the same rules may not apply for trademarking an AI-generated logo, a vast majority of small to medium businesses won’t pay the steep fees required to go through the trademarking process, leaving themselves with zero protection.

Where does AI artwork come from?
Generally, AI learning models are fed by imagery from the internet – whether that imagery was found in public directories, downloaded from paid or pirated sources, or bought and paid for. AI models learn from things that already exist.
Think about where that imagery came from…someone created it. And it’s nearly impossible to know who that someone is, and whether they were 1) fairly compensated for their contribution to the learning model or 2) whether they opted in to having their work(s) train the AI models.
Cite Your Sources
Have you ever seen an AI model that has cited their sources? If we the humans were to create a composite of things for something of value, and we didn’t cite our sources, we’d be accused of plagiarism and generally discredited. So why isn’t AI treated the same?
When you ask AI to provide sources, it throws a few dozen links in there that it may or may not have leveraged to any great extent, and it’s up to you to decide whether you’re going to trust it. There’s even a disclaimer on the bottom that says, “sometimes [the AI model] gets facts wrong so make sure to double check;” but do you?
How should AI be used?
I’m not naive; I know AI isn’t going anywhere. I’ve had to evolve my business to leverage it – it’s built into the tools I use every day, and speed has always been a factor in design and marketing. And yes, I know that as a business owner, you should always be looking for the next big thing that’s going to make running your business easier.
But AI should be a tool, not a replacement for real humans.
AI doesn’t possess empathy, and it doesn’t understand nuance. You can compare an AI prompt to writing a text message or an email that was ill-received when the intention was only positive; emotion doesn’t come through in text for most people. So how could AI “listen” to a business owner tell the story of their brand and pick up on the emotion and nuance that should be brought into the design of it? Or, know where to ask questions and get the deeper story that ultimately makes the brand?
Here’s how I use AI in my business:
- I use it for competitive research (and then go and verify what it’s showing me).
- I use it as a sounding board for content and ideas to see how they might work for my ideal audience, in addition to my circle of real people.
- I use it to check my tone on sensitive emails.
- I use Photoshop’s generative fill and expand features to remove filler, irrelevant background elements or to make a photo wider for a website (sparingly).
Here’s what I won’t use AI for:
- To write content for me – emails, social posts, blogs, newsletters, website content, etc.
- To create artwork – including patterns, icons, logos, or photos – of any kind
- For fact checking
When is using AI for logos okay?
In my opinion, never.
But what if I don’t have a budget? Use a word-only logo until you do.
But what if I don’t know a designer? Ask around, there’s a lot of us out there.
But what if I need a logo for something fast? You should probably slow down or you’re going to miss something crucial. Also, see the first point above.
Getting a logo via AI – whatever learning model or platform you use – is doing your business a disservice.
- AI doesn’t have the human touch; it can’t fine tune your logo to be right for your audience.
- It isn’t taking into account the larger strategy of using your logo.
- It’s not going to generate different usable variations for different scenarios.
And the worst part (from my perspective) is that even if you’re using an AI platform that can generate an SVG or EPS version of your logo, that file is going to be such a hot mess that you’ll have an incredibly difficult time using it again in the future. So if you start off with an AI logo to get something fast, with the intention of having a professional edit it later, you may find yourself having to start from scratch because a professional may not be able to (or may not be willing to) touch the AI generated items.
What’s Most Important?
In this era of speed (seemingly) being the most important thing, I’d challenge you to consider what you actually value most in the long run.
Is it doing the thing the fastest, saying “yep that will work,” and getting it out the door?
Or is it taking the time to do it right, and knowing you won’t have to revisit it later and spend more time (and potentially money) on it?




























