Why AI in branding is a bad idea

People really don’t like AI writing, especially when it’s used to sell them a product.

In a June 2026 WordPress VIP survey, 2/3 of consumers said they’re immediately turned off when they see brand messaging that looks AI.

The quick-and-dirty solution is obvious. Keep on using AI (it’s cheap) and tell consumers your messaging is done by humans. Voilà.

Well, not so fast. Researchers at Princeton University assessed this exact scenario and found that such subterfuge doesn’t work.$$Because the human brain is a very sophisticated pattern-recognition engine--which is a fancy way of saying we tend to know when we’re being lied to.

The Princeton researchers gave study participants identical pieces of text with various labels and their findings show that lying about the origin of your content just blows up in your face.

The researchers put a “human-written” label on AI text and the study participants caught the deception right away. They flagged it as manipulative and said they wouldn’t trust the brand behind it.

Funnily enough, when the researchers ran that same text by LLM judges, the AI models also flagged the content. Turns out, even machines don’t like writing by other machines.

LLMs are getting quicker and better but our ability to spot slop is improving just as fast.

Which means you can’t trick a buyer with a fake label. If you want people to believe your brand messaging is done by a human, then it must in fact be done by a human.