AI Creativity Tests: What They Mean for Small Business

AI now beats the average human on creativity tests. Here's what the new Scientific Reports study means for your small business marketing.

Quick Answer

A January 2026 study in Scientific Reports found that leading AI models now beat the average human on structured creativity tests. The most creative 10% of humans still outscore every AI model tested. For your small business, AI raises your creative floor. Your customer knowledge is still the ceiling.

I’ve watched AI get better at math, writing, and analysis for years. Creativity was supposed to be the moat. The one thing machines couldn’t touch.

A new peer-reviewed study says that moat is half drained. AI models like ChatGPT, Claude, and Gemini now beat the average human on standard creativity tests. Not someday. Now.

If you handle your own marketing, this matters more than you think. Here’s what the study found, what it means for your business, and how to stay ahead without panicking.

What Did the AI Creativity Study Actually Find?

The study found that some AI models score higher than the average human on divergent creativity tests. The most creative humans still beat every model. Researchers at the Université de Montréal compared more than 100,000 human participants against today’s leading AI models. The paper was published in Scientific Reports in January 2026. Professor Karim Jerbi led the team, with AI pioneer Yoshua Bengio among the co-authors.

The main test was the Divergent Association Task, or DAT. Created by study co-author Jay Olson, it asks you to list ten words as unrelated to each other as possible. It takes about three minutes, and it correlates strongly with broader creative ability.

Try it: a high-scoring answer looks like galaxy, fork, freedom, algae, harmonica, quantum, nostalgia, velvet, hurricane, photosynthesis. The further apart the meanings, the higher the score.

The team also tested creative writing: haiku, story synopses, and flash fiction, scored identically for humans and machines. The results split cleanly. GPT-4 and others beat the average human on the DAT. But the most creative half of humans outscored every model, and the gap widened among the top 10%.

“Our study shows that some AI systems based on large language models can now outperform average human creativity on well-defined tasks. But even the best AI systems still fall short of the levels reached by the most creative humans.”
Professor Karim Jerbi, Université de Montréal

What Does This Mean for Your Small Business?

It means the bar for average creative work just moved. Average is now available to everyone for $20 a month. Your competitors can produce better-than-average content, social posts, and ad copy without hiring anyone. The plumber down the street can generate polished social ideas. The bakery across town can draft email campaigns that used to need an agency.

I see it with my own clients. The ones who used to freeze at a blank page now ship content that sounds professional. They’re not winning awards. They’re just not losing customers to weak copy anymore.

If “good enough” creative was your edge, you don’t have an edge now.

Does This Mean AI Replaces Human Creativity?

No. AI raised the creative floor. The most creative humans still own the ceiling, and the study’s own data says so.

Think of spell check. Before it, clean spelling was an advantage. After it, everyone’s writing looked clean, and the advantage moved to people with better ideas. AI is becoming spell check for creative work.

So the winners won’t be the businesses trying to out-create the machines. They’ll be the ones pairing AI’s volume with judgment the machines don’t have.

What this looks like: say you run a landscaping company. You ask AI for twenty seasonal promotion ideas and get them in ten minutes. Nineteen are generic. You spot the one that fits what your customers actually ask for every spring. You sharpen it with details only you know, and run that. AI brought volume. You brought the judgment that made it work.

How Do You Use AI for Creative Work Without Losing Your Judgment?

Four rules keep AI useful and keep you in charge. I use these with every client.

1. Use AI for volume, not decisions

Let it generate twenty concepts. You pick the one worth refining. Generating is cheap now. Choosing well is the skill.

2. Draft with AI, edit with judgment

First drafts of emails, landing pages, and posts are AI work. The edit, where you apply what works in your market, is yours.

3. Feed it context

Specific detail about your customers, offers, and voice produces specific output. Generic prompts produce generic results. This is most of the gap between owners who love AI and owners who gave up on it.

4. Protect your strategy

AI generates options. Only you know your goals, margins, and capacity. The decision layer never gets delegated.

Want Your Team Using AI Like This?

Chad and I run hands-on AI training workshops for teams, chambers, and owner groups. Your people learn to get useful output from these tools, with your business as the example. No theory lectures.

See How the Workshops Work

What Can’t AI Do for Your Business?

AI can’t know your customers, and that’s where small businesses actually win. The study measured structured creativity tests. Running a business takes more than generating clever ideas. You need to know which clever ideas drive sales and which just sound good in a meeting.

AI can brainstorm menu names for a restaurant. Only the owner knows their crowd wants “Great Burgers,” not “Artisanal Grass-Fed Beef Medallions.”

The most successful owners I know don’t win on raw creativity anyway. They win by understanding their customers better than anyone, executing consistently, and solving real problems. AI can make you more creative. It can’t make you care more about your customers than your competition does. That part’s still yours.

What I’m Telling My Clients

I’m not sugarcoating it. The creative bar moved. If your marketing was already weak, AI just made everyone else’s stronger.

I’m also not sounding an alarm. The advantage is shifting to people who combine AI volume with deep knowledge of their market. That’s the one game small business owners are built to win. You know your customers, your margins, and which promises you can keep. AI doesn’t.

Start experimenting now. Pick one creative task you dread, run it through the four rules above, and compare the results. Want AI doing more than marketing? My guide to AI automations for small business covers the eight setups that pay off first.

Not Sure Where AI Fits in Your Business?

That’s the whole point of a discovery call. Twenty minutes, plain English, and you’ll leave knowing your smallest useful next step, whether you work with us or not.

Book Your Free Discovery Call

Frequently Asked Questions About AI and Creativity

Which AI models were tested in the creativity study?

The researchers evaluated leading large language models, including GPT-4-class systems, against more than 100,000 human participants. Testing covered the Divergent Association Task and creative writing tasks like haiku and flash fiction, scored identically for both groups.

What is the Divergent Association Task?

The Divergent Association Task (DAT) is a short psychological test that measures divergent creativity. You list ten words that are as unrelated to each other as possible. The semantic distance between them determines your score. It was created by researcher Jay Olson and correlates strongly with broader creative ability.

Does this mean AI will replace human marketers?

No. AI now beats the average human on structured creativity tests. The top 10% of creative humans still outscore every model tested. Strategy, customer understanding, and brand judgment still require a human. AI raises the floor for creative output rather than replacing the people directing it.

How can a small business use AI for creativity without losing its voice?

Feed the AI detailed context about your brand, customers, and goals, then use it for volume and first drafts. Edit everything with your own judgment before it ships. Specific prompts produce output that sounds like you. Generic prompts produce generic marketing copy.

Is the AI creativity study credible?

Yes. It was published in Scientific Reports, a peer-reviewed Nature Portfolio journal, in January 2026. The research team at Université de Montréal included Professor Karim Jerbi and deep learning pioneer Yoshua Bengio. The human sample exceeded 100,000 participants.

Neil Gass headshot

Neil Gass

Neil helps small business owners cut through AI hype and build systems that actually work. If you’re tired of chasing trends and want a practical plan, he’s happy to talk.

Book a Call