You read a post that makes you laugh, teaches you something useful, or gives you a genuinely good idea. Then you find out it was written by AI.
Would you like it less?
For many people, the answer is yes. But there is a catch: readers are pretty bad at identifying AI-generated content in the first place. That puts brands in a dilemma: a piece of content can seem perfectly acceptable until its origin becomes known.
The reaction also depends on the kind of content involved. Knowing that AI helped write a two-sentence product description may mean very little to a customer. A fake testimonial is a different matter entirely.
Can People Even Tell When Content Is AI-Generated?
A study of AI-generated images collected roughly 287,000 evaluations from more than 12,500 participants. The overall identification rate was 62%. That is better than random guessing, but nowhere near reliable. Participants had particular difficulty with some landscapes and urban scenes.
Text is hard to identify too. A study of short stories found that participants were generally unable to distinguish human-written stories from AI-written ones. The AI-generated stories received high ratings for quality and engagement. Participants also rated stories more highly when they believed a human had written them, regardless of the actual source.
That creates an important distinction. There is the quality of the content itself, and there is the assumption we make about who or what produced it.
Those two things can influence each other, even when the content stays exactly the same.
Does Knowing About AI Change Your Opinion?
A 2025 experiment found something that makes the AI-content debate more complicated. Participants read answers written by humans and AI. Some knew who had produced each answer; others did not. When the source was hidden, participants tended to prefer the AI-generated answers. After they learned that an answer came from AI, their preference changed.
A separate study, which examined nearly 1,000 evaluations across six types of writing, found that disclosure of AI involvement changed how the content was rated. The effect was stronger for social and interpersonal writing, where the perceived source of the words mattered more to the evaluation.
That tells us something. The same piece of content can receive a different reaction depending on what the audience knows about its creation.
But the effect is not the same everywhere.
The Type of Content Makes a Difference
AI-generated content covers a lot of different things, and audiences do not necessarily judge all of them by the same standard. The easiest way to understand the difference is to look at what the content is actually claiming.
Product Information
A product description makes a straightforward promise: the information should be accurate.
If AI takes real product specifications and turns them into a social media caption, there is little reason for the audience to care how the sentence was written. They want to know what the product is, what it does, and why it might be useful to them.
The same applies to a short post explaining a product feature, a recipe, or a basic how-to guide. The important part is that the underlying information is correct.
Creative Advertising
Advertising gives brands much more freedom.
Take a pair of sunglasses. A brand might create an image of a man and woman walking through a sunny city, stopping at a café while wearing those sunglasses. The people and the setting can be fictional while the product itself is real.
That is a creative representation of how the product could look in everyday life. The image does not claim that this exact couple exists or that the photograph documents a real afternoon.
The same principle applies to a caption. Something like “Sunny afternoons call for good company and your favorite shades” is a piece of advertising copy. It creates a mood around the product without claiming that a specific event actually happened.
AI can be useful here because a brand may need many different creative ideas for the same product. A pair of sunglasses can appear in a city scene, at the beach, on a road trip, or at an outdoor café. Each scenario can communicate something different about the product without pretending to document reality.
Personal Experience
The situation changes when the content claims that a real person experienced something.
A founder might write about leaving a job, developing the first version of a product, making an expensive mistake, or nearly closing the business. Those details have value because they describe something that actually happened.
AI can help turn the founder’s notes into a polished story. It can also help organize a rough draft or suggest a better way to explain the experience.
Inventing the experience is different. If the events never happened, presenting them as a personal story gives the audience a false impression.
Testimonials And Reviews
Customer opinions fall into the same category.
A company can use AI to shorten a genuine review, correct its grammar, or turn a longer comment into a social media post. The underlying experience still belongs to a real customer.
Creating a fictional customer and giving that person a made-up opinion is a different matter. The issue has nothing to do with AI writing a sentence. The issue is presenting an invented endorsement as a genuine one.
The distinction is easy to apply: fictional advertising can create a scenario, but it should not pretend that the scenario is evidence.
Expert And Factual Claims
Expert content has another set of expectations.
A doctor, lawyer, accountant, photographer, or other specialist might use AI to turn their own notes into a social media post. The professional can check the claims and decide what information is appropriate.
The situation changes when AI generates advice and a professional identity is attached to it without genuine expertise behind the content.
The same principle applies to statistics, research findings, product claims, and other factual statements. AI can help communicate the information, but the information itself needs a real basis.
The Important Distinction
The question is therefore not just whether AI created the content. It is what the content asks the audience to believe.
A fictional couple wearing real sunglasses is creative advertising. A fictional customer praising those sunglasses as if they had actually bought them is a fabricated testimonial.
Those two pieces could both be generated by the same AI system. The difference lies in the claim being made.
That distinction matters because AI-generated advertising does not have to imitate reality to be effective. A fictional scene can communicate a product’s style, setting, or use without pretending that a real event took place.
The line becomes important when creative material is presented as evidence of something real.
The Bigger Problem Is Poor Content
There is plenty of evidence that people are becoming tired of low-quality material produced with generative AI.
A 2026 Gartner survey of 307 U.S. consumers found that 49% believed generative AI had made the quality of available content worse. Among Gen Z and millennials, the figure rose to 57%.
Those numbers are worth taking seriously. They do not mean people reject every piece of AI-assisted content. They suggest that people are becoming more cautious about what they consume online.
You can see why in everyday social media.
A weak AI-generated post might contain:
- a vague statement that could belong to almost any company
- five sentences that say very little
- an exaggerated claim with no evidence
- a product image that looks polished but makes no sense
- the same motivational language used by hundreds of other accounts
The problem is the lack of a good reason to spend time with the post.
Give an AI system useful product information, a clear idea, real facts, and a human point of view, and the result can be very different.
So, Do People Really Care?
Some do. But “Is this AI-generated?” is not the whole story.
People tend to care much more when AI use changes what they believe about the content. If a post claims personal experience, readers may want a real person behind it. If an image claims to document something that happened, authenticity becomes crucial. If a product description gives accurate information, the production method may be far less significant.
What matters in the end is what the content says about the business. Someone still needs to decide what is accurate, relevant, and representative of the brand.
The goal is not to make content look human so nobody notices the AI. It is to make content worth reading in the first place.
For a brand that publishes frequently, AI can help with the volume of creative production: generating visual ideas, drafting captions, adapting concepts into different formats, or preparing campaign material.
AI tools like Stryng can use information from a webshop to create social posts around specific products and creative scenarios, with the brand making the final call on what gets published.
If readers get something useful, accurate, original, or interesting, the question of how it was produced becomes much less important.
That may be the real lesson behind the debate over AI content.
