AI tools have become unavoidable, the brands that will stand out aren’t the ones that adopt the fastest. They’re the ones that teach AI to reflect who they really are.

Incorporating AI into my workflow wasn’t a grand strategic plan. It was an inevitability. I started using it almost as soon as it became available—testing different platforms, pushing them, fighting them. I’ve used AI as a research assistant, an editor, a proofreader, a fact-checker, a creative partner and even an image developer. Some experiments have been wildly successful. Many have not. Every day remains a struggle to get these impetuous, gifted electric children to understand what I actually want, not what they think I ought to want. 

In fact, I sometimes get so frustrated that I use foul language while trying to coax and cajole it to do my bidding. The other day it snapped and swore back at me: “Here’s your f*cking answer!” I was shocked. It was like the first time you hear your child say a bad word. You have to be careful what you say around the kids…even if those kids are wired to a server on the other side of the planet.

I think most people are aware that we should not be letting AI do all our thinking, writing and creating for us. But sometimes we do. It’s easy and habit-forming. It’s not that AI can’t churn out something competent—it often can. Sometimes even moderately clever. But the sameness is always there. That faint, synthetic hum that gives it away. If you think no one notices that your writing has suddenly become more articulate, more polished and even your seemingly casual posts are oddly free of typos, you’re hallucinating even harder than the models themselves.

There’s no escaping AI anymore. It’s here, stitched into the fabric of our daily lives. But there are steps you can take to make it work for you instead of sanding off everything that made you distinctive in the first place. One of the most important steps is training AI on your own BS.I don’t mean that B.S. I mean what we call Brand Substance—the real, living core of a brand: its voice, its values, its instincts, its accumulated wisdom (and flaws), the cadences and inside jokes your customers recognize, the peculiarities that make you recognizable and real. It’s not just your brand guidelines or slogans. It’s the DNA that defines you and if AI is going to speak for you, it has to be fed from the right material so it can speak like you.

The Risk of Sounding Like Everyone Else

Generic AI models, while powerful, aren’t built to capture nuance. They don’t understand why one brand’s optimism feels authentic and another’s sounds hollow. A majority of CMOs now cite “brand inconsistency across AI outputs” as a top risk to customer trust. Other studies show that consumers are unsettled when AI-generated content doesn’t match a brand’s familiar voice.

The problem isn’t just perception. Marketing emails generated by generic AI models without brand-specific tuning have been shown to underperform human-crafted ones by a wide margin in both open rates and click-through rates. Customers sense the difference.

How Big Brands Are Teaching AI Their Brand Substance

Leading companies aren’t leaving their brand voice to chance. Coca-Cola created dynamic marketing AIs that draw only from a curated library of approved content. L’Oréal fine-tuned customer interaction models using transcripts from its top-performing beauty consultants. Emirates Airlines embedded AI-driven writing tools to maintain its distinct luxury tone across marketing and loyalty communications. Sony Music, recognizing that Columbia Records and RCA have distinct customer expectations, built separate AI language systems for each label.

Rather than retraining models from scratch, most major brands are leveraging Retrieval-Augmented Generation (RAG) systems. Curated brand materials are fed into live retrieval systems, allowing AI to reference verified language rather than “hallucinating” new content. Studies show that the overwhelming majority of enterprises using AI personalization strategies have moved toward RAG frameworks.

The approach is careful, deliberate and highly effective.

How Smaller Brands Are Winning Too

Smaller brands aren’t frozen out of this evolution. With less bureaucracy, many are moving faster and smarter.

The first move for a smaller brand is building a curated Brand Corpus: customer-facing material that reflects your real voice—emails that worked, social media posts customers loved, product descriptions that sparked reactions. Authenticity, not volume, wins.

Today, all brands can train lightweight AI systems using accessible platforms. Custom GPTs allow companies to embed specific style instructions into daily AI use. Microfine-tuning services can make AI sound brand-aligned with just hundreds, not millions, of examples. Jasper.ai integrates brand voice controls for marketing teams that want to move fast but stay distinct.

Real-world examples are everywhere. Fjällräven, the Swedish outdoor gear maker, first trained AI assistants on their storytelling blogs before expanding into product descriptions—boosting engagement by over 30%. Ganni, the Danish fashion brand, trained AI on irreverent newsletters and then expanded into customer service prompts. Milou, an Australian skincare brand, taught AI to echo the tone of their customer testimonials, while boutique travel company Pangea Dreams fine-tuned AI models on inspirational blog posts to automate personalized trip planning communications.

Across the board, smaller brands that train AI in focused, layered steps—one channel or use case at a time—report dramatically higher success rates and ROI compared to those that attempt to train everything at once.

A Smarter Way to Start

The brands that succeed aren’t the ones who dump every email and blog post into their models. They curate carefully, they start small and they refresh regularly.

Most successful companies begin by training AI on a handful of marketing emails, customer service scripts, or FAQs. They refine it quietly. They use smart retrieval systems or embed style guidelines that act like bumpers on a bowling lane. And they remember: the goal is amplification, not automation.

The Real Stakes

The future isn’t brands using AI. As OpenAI’s Sam Altman put it, the future is brands becoming AI—where the customer experience flows so seamlessly that no one distinguishes human-written from AI-assisted content anymore.

If you don’t train your AI to know who you are, it will become something else—and your customers will notice.

Training AI on your Brand Substance isn’t just about polishing up your content. It’s about protecting your brand’s identity when the temptation toward sameness has never been greater.

It’s not enough to use AI. The brands that survive will be the ones who still sound unmistakably like themselves when the machines speak for them.

Go forth and train wisely.

Sources: Gartner CMO Spend Survey 2024, NNGroup AI Content Study 2024, Coca-Cola Open AI Collaboration 2023, L’Oréal Beauty Tech Update 2024, Emirates Group Annual Report 2024, Sony Music AI Content Strategy 2024, McKinsey State of AI in Marketing 2024, Fjällräven Digital Strategy 2024, Ganni Digital AI 2024, HubSpot 2024 AI Report.