AI-pilled bosses and clients are turning designers into ‘pixel pushers’

Emily Foran, a creative director and graphic designer, has not been asked by clients to come up with an original idea from scratch in a year. 

“That’s kind of killing my soul,” Foran says. “I’m having clients, new and existing, send me crap they had Claude or ChatGPT mock up for them, and then ask me to either clean it up or straight up work with it. And I hate that. It’s such an insult.” 

These days, she feels more like a “pixel pusher” than a creative, a concern echoed by many of her peers. Foran, the president of the American Institute of Graphic Arts (AIGA) Maine, is one of more than a dozen graphic designers Fast Company interviewed who say that generative artificial intelligence has completely changed their relationship with their bosses, clients, and work itself over the past 18 months. 

Clients have asked for lower rates, because they claim they’re now the ones coming up with images alongside an AI assist. Some designers have been laid off and explicitly told that AI would replace them. Others are only getting guidance from their bosses that amounts to “Ask ChatGPT.” Many still have jobs, yet feel like their work and perspective are undervalued or seen as replaceable. 

Foran has put her foot down; she has an AI policy on the website for her studio, Foran Concepts, that asks clients not to upload her work into generative AI systems or use it to generate feedback for her. Most of the designers Fast Company interviewed don’t view themselves as dogmatically anti-AI, but they’re wary of how the technology is affecting their industry. Overall, there’s a sense that ChatGPT, Claude, and other generative AI systems now get to do the most fulfilling part of the work, while humans are left with the tedious aftermath. 

‘I’VE ALREADY CREATED IT’

Amanda, a graphic designer in Canada, initially felt optimistic about the potential for AI images in the industry. 

“A lot of people have a hard time articulating what they’re looking for, and now here’s a tool that just shows it,” she says. She thought it might help her clients share their vision for the project, which she could use as inspiration for the creative part of the process. But in her experience, that isn’t how the technology has played out. Not even close.

“They always say, ‘Make it look more professional, and then get it ready for the printer,’” Amanda says. Then, they add, “By the way, now that I’ve saved you the trouble of developing the creative, it should be cheaper, right?”

But using AI hasn’t led to any cost savings on the design side. In fact, designers say that images rendered by AI ultimately result in more work for them on the back end. 

One graphic designer at a small firm based in Boston, who asked to remain anonymous to discuss her workplace candidly, says that AI might be streamlining the process for whoever is prompting it, “but for me, it’s definitely not cutting down my time. What used to take me an hour to do might be taking me three hours, because I have to collapse 20 bullet points that are in 20 different boxes into one text box.” 

The formatting on AI-made assets is wacky, and the sizing is all wrong. “When I sit down and I open up a project and I can tell that Claude built it, like, okay, there goes the rest of my afternoon,” she says. Still, her company leadership is pushing the use of AI in their daily work and she sees it from clients too. “It’s just coming from all sides,” she says. 

Many clients have bought into the hype that AI will make things faster, and are trying to set timelines accordingly with designers. “We have had clients come to us and say, ‘Well, since we gave you the concept, this shouldn’t take that long,” says Michele Kasl, owner and creative director of Premier Designs. “I have never heard that more in my life than I have the past six months.”

Kasl sees it as part of the devaluation of design work. “Just because AI can do something doesn’t mean it should, or even that the result it will produce is good or good enough,” Kasl says. Clients are “dabbling in areas where they lack the expertise, but they feel like they’ve gotten the expertise because of the output that AI has put in front of them.”

So what happens when clients ask Premier Designs for a truncated timeline because they’ve generated their own assets? “Frankly, we’ve lost business because of it,” Kasl says. 

Robert Lodi, whose company Rock Pixel Scissors builds and fixes websites for companies of varying sizes, says this year he has seen “an influx of AI-built mockups, designs. Some people think they’re ready to go and we look at them and at a glance, we’re like, ‘all right, this isn’t bad.’ . . . But once you start looking more, like with a lot of AI content, you start to see some issues.”

AI-generated handoffs are happening so often that Lodi has added a redesign phase to the company’s workflow, “where we try to take what they’ve sent and work with them to make it something that’s functional and usable.” One major issue he sees? None of the AI-made websites are designed for accessibility. “I have yet to see one that’s going to pass an accessibility test,” he says. “And that’s important because there is now a thing called accessibility lawsuits.”

Over the summer, he worked with a nonprofit to design an email newsletter. The client gave him AI-generated mockups that were simply unusable—they were way too big. “If you tried to cram it into an email, it’s never going to display well or it’s going to take 17 times their budget to actually make it work and it’s going to have to be custom code that they can never edit or reproduce in their email system,” Lodi says. 

Despite these clear issues, the client insisted on seeing his mockup formatted for the email system. “We’re getting more of that now because they’re happy with their things, they’ve put work into their prompts, but they don’t realize that it’s not ready for the real world,” Lodi says. 

Managing clients’ or bosses’ expectations and figuring out graceful ways to tell them their ideas aren’t doable has long been something that designers need to navigate in their jobs. But even though it isn’t a new problem, it has been exacerbated by AI. “There’s a disconnect,” Lodi says. “They think they have finished products when really, they’re further behind than if we had just done it right from the start.”

He even told one client that “if you come back with a bunch of AI stuff, then the price is going up,” Lodi says. The client laughed when he said that, and then returned with more AI-generated mockups. “The price did go up,” he says.

AN AVERAGE DESIGN BY DEFINITION

When clients decide to hire designers, they’re spending money to get something original. By creating mockups with generative AI, though, they’re effectively ensuring the opposite. “I did try to have conversations with either prospects or clients to say, ‘You realize that AI is the average of everything that already exists out there?’” says Amanda, the graphic designer in Canada. “Do you want to look like the average sum of your 30 competitors?”

For some clients, the answer seems to be “sure.” 

“It’s like, if it’s good enough, they’re happy with it,” says Foran, the creative director in Maine. And certainly, there are many businesses adopting AI that were never going to pay a designer to help them. But Foran has been disheartened to see some of her long-term clients hop on the AI train, too. 

“On one hand, I’m grateful because I can still pay my bills, support my family with my job,” Foran says. “On the other hand, where did the fun go? Where did the innovation go? Where did the art of it go?”

Kasl thinks that, with a tricky economy, businesses with a budget crunch see design as one of the first line items on the chopping block.

“There’s no immediate consequence to bad creative,” Kasl says. “The fallout isn’t immediate. No one ends up in the hospital because of a bad flyer design, or no one ends up in jail because they put together an ugly postcard. . . . Brand perception is going to be affected by these choices that they’re making.”

Erik, a marketing designer, experienced this when he worked for a small hobby company where AI kept getting more deeply embedded in the design process. Shortly after he started working there, his employer set him up with a corporate ChatGPT account to use for quick iterations to use in-house. “It made sense at that level,” he says. 

But then, the CEO began asking for ever-quicker turnaround times to make up for a lack of planning. “We have a Memorial Day sale coming up in a week. Nothing gets talked about. Nothing is discussed. We have ideas, we pitch them, it’s put on the back burner,” he says. Then, mere days before the event, leadership wanted to move forward with “kitschy ideas.” When the design team brought up that there wasn’t enough time to go through a proper design process, Erik says, higher-ups said that, “We’ll just throw it in ChatGPT.”

When he was tasked with new projects, he’d try to get insight from his direct boss about the messaging or the look they wanted. The constant answer? “Just ask ChatGPT.” And his manager would ignore Erik’s decade-plus of design experience in favor of whatever the AI model spit out. Ultimately, the company got called out for using “AI slop” in its Facebook and Instagram ads. Erik eventually lost his job because he kept clashing with others over AI use. 

TROUBLE FOR THE NEXT GENERATION OF DESIGNERS

A designer who spent the last half decade working in-house for tech startups began to hear AI as a major buzzword over the past few years. (She asked to remain anonymous because she is searching for work.) She and her product design team told higher-ups they would use AI for efficiency and workflows, but not to generate ideas. That pushback led to internal tensions. 

During one product launch, she recalls that leadership tried to show how much quicker they could be by creating a landing page using AI without any help from designers. “I said, ‘But your information hierarchy is all over the place. The product animation looks like any other product.’” Plus, the landing page included information about the product that wasn’t accurate. Ultimately, that version did not go live.

Still, she kept getting feedback that the team needed to go faster, but never got a sense of what they were supposed to do with the time they were allegedly gaining. “You’re just outputting a lot of noise,” she says. “It all looks exactly the same.” She advocated for building out a brand asset library using original photography and video shoots to distinguish the look of the company. That idea did not fly. “I was encouraged to just generate video from AI,” she says.

The startup decided to cut out the middle man: her. “Eventually, the company leaned into using AI and a lot of roles were eliminated,” she says, including hers. “Which is sad to see because all the work that I had done to make that brand differentiate and lead with human-first storytelling—now it just looks exactly the same as with all the other competitors.”

Alan Peters, a designer and brand builder with Jupiter Visual, has a similar story. He worked as the lead graphic designer at an online sticker company before the business laid off the creative team in marketing in favor of an online AI service. But the product isn’t better, he says; it’s just cheaper. 

“It’s not doing a better job than a designer’s going to do for creating an asset from scratch,” Peters says. “It’s just completely generic. It references the colors in a vague way but it misses all the marks on typography, misses all the marks on representing the products properly, on brand voice and then just the design quality is very poor.”

In the lead-up to the layoff, Peters noticed the company’s higher-ups gravitating towards using AI for design tasks. His team used it too, but mostly “to make workflows more efficient.” For example, while Peters pushed for actual product photography, he was fine using AI in post-production to do things like adding a background. But Peters noticed a gradual shift in the way company leaders approached the back-and-forth of a design process.  As he sees it, the willingness to collaborate changed.

“There was a lot of feedback when it was the human designers. . . . You might see rounds of revisions on assets to get where the marketing team wanted something to be,” he says. “But then with the generative AI, it’s like, ‘Well, okay, we prompted it. Ship it.’”

As bosses and clients require designers to use AI for more of their job, those decisions ripple through the industry. It’s leading some designers to leave the industry altogether. Foran has observed this through her work with AIGA Maine, “especially [among] younger people who barely even got their feet in the ground,” she says. “It’s like they don’t see the possibility because there haven’t been opportunities.”

That’s a concern for Amanda, too, who says that there’s something lost when AI will do the work that traditionally fell to younger designers, like business cards, postcards, and flyers. “That’s where you get your reps in, right? You get your 10,000 hours doing these smaller things and then eventually move up to the more strategy work,” she says. “As an experienced creative, we’re still being hired for strategy work, whereas a junior can now easily be replaced by AI to turn out those sort of smaller things.”

While some companies will continue to place an emphasis on good design, the rise of the AI-abetted armchair design boss amounts to a great unmasking. Suddenly, designers see more clearly than ever before who values their work and who is willing to kick them to the curb or ignore their insights as soon as they’re able.