AI Content Extracts Brand Personality and Creativity

Many of us are using ChatGPT, Claude, Gemini, Grok or Kimi to write content, and we’re sharing more articles and social media posts than ever.

Of course, AI platforms save us time because they are good at producing competent, readable content, as long as our prompts are good. However, they are not good at injecting a brand’s personality and secret sauce into articles, which are the important creative ingredients that make our brands genuinely different.

“This matters for brands and businesses because we operate in highly competitive markets and have to differentiate ourselves on reputation, personality and distinctiveness, and not just price,” says Christopher Lamotte from Real Marketing Transformation who is sharing this article.

I’ve been seeing this bland ‘McDonaldization’ in first draft articles like email newsletters, press releases, brochures and website content sent to me by clients for editing. That’s why I always recommend to my clients that they should edit content created by AI themselves before to inject their brand and personality back into it (as well as re-prompting).

Recent academic research supports this major issue. Importantly, it does not suggest that AI writing is always poor. It reveals that large language models (LLMs) can improve the quality of individual pieces of writing, particularly if you find writing hard work, like many of us!

“By the way, the top three AI platforms that are currently the best, or ‘least worst’, at preserving brand personality and selling messages are understood to be Claude, ChatGPT and Gemini, in order of priority. Caveat: yes, AI is becoming more effective so it may get better at communicating brand personality.”

However, LLMs generate content that is less distinctive, more average and more likely to be boring. It also means that they don’t communicate well.

According to a study in Nature Human Behaviour, 2026, AI writing reduces the diversity of language by 21-50%. Researchers examined more than 880,000 texts across seven datasets and three studies, investigating how LLM writing assistance changes the language people use.

In other research, they analysed 2,200 student-written college admissions essays compared to GPT-4 written essays – they found that human writing contributed 2 to 8 times more new ideas than AI writing (Computers in Human Behaviour: Artificial Humans, 2025, by Moon, Green and Kushlev).

The remainder of this article has been influenced by helpful research presented in Matthew Hutson’s article ‘Bland new world: is AI making us all think the same?’

Researchers are putting their anecdotal observations to the test by exploring how generative AI (genAI systems that create text, images, etc) is homogenizing both culture and thought. In a March paper, Zhivar Sourati (a PhD student at the University of Southern California) and his co-authors noted that this phenomenon is similar to ‘McDonaldization’, which has led to more uniformity, echoing process-driven, fast-food efficiency.

Studies suggest that AI tools can flatten language use, creativity and cultural values, and there is evidence that the technology can even influence the decisions we make, the way we act and the opinions we hold.

Emily Wenger, a computer scientist at Duke University in Durham, North Carolina, says the risk of homogenization is serious:

“If we’re all using AI to write our emails or whatever, what do we become as a species? We have suffered as a society when we silence or ignore edge voices.”

To test the creative homogeneity of AI models, Wenger and a colleague tested 22 LLMs and 102 people on three creative tasks – one included divergent thinking that asked the models and participants to name alternative uses for common objects. They found that LLMs’ responses were more similar to each other than human responses.

Other researchers mapped the narrative features of short stories written by people and five LLMs by analysing the plots, characters, settings and other elements. They found that the AI-generated stories clustered together on the basis of their narrative features, but that those written by real people were more varied.

In summary, good marketing content should not rely on unedited AI generated content. In my opinion supported by research, all LLM written content must be edited with the brand personality, tone of voice, differentiators and selling messages put back into any articles before they are published.

The second part of this article will be published within the next two weeks, so watch out for it!

Contact Christopher Lamotte at Real Marketing Transformation for an informal Zoom or call to discuss your content and marketing challenges.