Language Dropdown

Most Consumers Can Spot AI-Generated Content

Most Consumers Can Spot AI-Generated Content — Here’s What Marketers Should Know

New studies show that people are getting really good at noticing when content is made by AI—whether it’s a photo, a painting, or even an article.

If you’re a marketer or content creator, this could change how you use AI in your strategy.

Let’s break it down 👇

📸 Study #1: People Can Spot AI Images Easily

A study by digital marketing expert Joe Youngblood asked over 4,000 people in the U.S. to look at real photos next to AI-generated ones. On average, 71.6% of people correctly picked out the AI image.

Some types of images were even easier to detect:

  • 👩‍🎤 Scarlett Johansson as Black Widow (celebrity): 88.8%
  • 🌄 Italian countryside (nature): 88.5%
  • 🐣 Baby peacock (animals): 88.0%
  • 🪐 Jupiter (space): 83.6%

But not all AI images were obvious:

  • 🗼 Eiffel Tower: only 18.1% spotted the fake
  • 🎨 George Washington painting: 50.9%

So, even though AI is getting better, most people can still tell what’s real and what’s not.

✍️ Study #2: AI-Written Content Is Also Easy to Spot

A separate study by Hookline& asked 1,000 Americans about AI-written content. The results?

  • 🧠 82.1% said they can detect AI-written text at least some of the time.
  • 👶 Among younger adults (22–34), that number jumps to 88.4%.
  • 🚫 Only 11.6% said they never notice AI content.

Hookline& CEO Christopher Walsh Sinka summed it up:

“Writers and brands aren’t sneaking AI-generated content past readers.”

🚨 Why This Matters for Brands

If people can easily spot AI content, what happens to your brand’s image?

According to the studies:

  • 50.1% of people think less of writers who use AI.
  • 40.4% view brands more negatively when they use AI content.
  • Only 10.1% would view those brands more positively.

📉 Older consumers (ages 45–65) were especially critical. Nearly 1 in 3 said they don’t like AI-written content at all.

When Is AI OK to Use?

It’s not all bad news. People are more open to AI being used behind the scenes for things like:

  • 🧠 Brainstorming (53.7%)
  • 🔍 Research (55.8%)
  • ✂️ Editing (50.8%)
  • 📊 Data analysis (50.1%)

In the image study, people said it’s fine to use AI for fun stuff like:

  • Memes
  • Cartoons
  • Diagrams
  • Video game artwork

But for serious content (like ads or blog posts), they still prefer human-made work.

🔍 What Should Marketers Do Now?

If you’re using AI for content, here’s how to move forward smartly:

  1. Be Honest: Most people can spot AI anyway, so be transparent about using it.
  2. Put People First: Let AI help with research and editing, but keep humans in charge of creative work.
  3. Focus on Quality: Real, thoughtful content still builds more trust.
  4. Know Your Audience: Younger users are more chill about AI—older users, not so much.

👇 TL;DR

  • 🧠 Most consumers can tell when content is made by AI.
  • 📉 AI content may hurt your brand image—especially with older audiences.
  • ✅ People are OK with AI for research, editing, and memes.
  • 🧑‍💻 Let humans handle the creative side to keep your brand trusted.

Bottom Line:
AI is a powerful tool, but people know when you use it—and they care. Use it smartly to support your work, not replace your creativity.

Facebook
LinkedIn
Threads
Reddit
Telegram
X
WhatsApp
Picture of Kumail Mehdi

Kumail Mehdi

I am a goal-driven person, I work well with people and like to challenge myself in different ways. I also want to have a great career that can develop me as an individual and an employer as well, so as to be part of a positive working environment where I can learn and grow. My interests include reading, swimming, and going out for fun.

Related Posts