AI is about to expose how strong your brand really is

Most companies are asking how AI can help them create more content, personalize campaigns or make customer service more efficient. All useful questions, but I think they miss a much bigger shift.

AI is moving from generating things for brands to making decisions on their behalf. As agents become more autonomous, they will decide what to recommend, when to sell, when to hold back and how to respond to a customer. Eventually, they may even decide what kind of experience or service to create for that individual.

Which raises an uncomfortable question: does your brand actually provide enough direction to make those decisions?

For years, the test of brand consistency has largely been: does this look and sound like us? Increasingly, we'll need to ask: did it make a choice that we would make? Consistency of expression is becoming consistency of judgement.

A tone of voice is not a point of view

Most brand platforms were never designed to answer that second question. They talk about being authentic, progressive, empowering, human or customer-centric. Those words can sound convincing in a presentation. They become less helpful when something has to make a choice.

Should it push or pause? Simplify or explain? Sell something or decide that, actually, this customer doesn't need anything right now? And what happens when the customer's best interest conflicts with a commercial target?

Experienced people have been filling in those gaps for years. A designer, a service employee or a salesperson brings context, experience and an understanding of the company that never quite makes it into the brand book. They know when to bend a rule, when to ask someone and when a technically acceptable answer would feel completely wrong.

An AI agent will interpret vague instructions too. But we shouldn't assume it will fill the gaps in the same way. As we hand more decisions to systems, we expose how much of the brand lives in people's heads rather than in the strategy itself.

If that thinking doesn't find its way into the system, the objectives it has been given will carry more weight. Engagement, conversion, efficiency or revenue may be perfectly clear. What the company means by “customer-centric” may not be.

Your AI needs more than a personality

Imagine a wellness brand whose agent has access, with permission, to a customer's sleep data, exercise patterns, calendar and purchase history. It could help create a personalized recovery program and recommend products when they actually contribute to the customer's goal.

But what happens when the customer doesn't need another product?

Within the same evidence and safety limits, a performance brand might help them structure their routine around a training goal. A brand centered around balance might help them leave more space in their calendar. Both could use the same technology and data, yet make different choices about what to prioritize. And both need to recognize when a question requires a professional.

If our imaginary wellness agent has to work a product into every solution, we're looking at advertising with a very sophisticated personalized interface. Sometimes a credible agent may have to say: you don't need to buy anything. Whether a company allows that tells us quite a lot about what the brand really stands for.

As more companies gain access to similar AI capabilities, those choices become a bigger source of differentiation. The colors, language, and interface will still matter. But customers will also experience a brand through what its agent recommends, what it prioritizes and what it refuses to do.

Brand strategy is moving into the product

Of course, brand shouldn't determine every decision an AI system makes. Product policy, legal requirements, risk, governance and commercial objectives all matter. Brand has to become an input into those decisions, contributing a clear point of view about the relationship the company wants to have with its customers.

Saying “we put the customer first” doesn't settle a difficult trade-off. Deciding when you'll accept a smaller sale to help someone make a better choice gets closer. Product, commercial and risk teams then need to agree how that works in practice, what the agent may do and how they'll know whether it's doing it well.

This has an organizational consequence. Brand needs to be in the room when those systems are designed, when their objectives are set and when their behaviour is tested. Bringing marketing in at the end to make the agent sound like the company leaves the most consequential brand decisions to everyone else.

For the wellness brand, that means agreeing whether a useful recommendation with no purchase counts as success. If the brand team talks about wellbeing while the product team measures only conversion, a warmer tone of voice won't resolve the contradiction.

We need clearer answers to a few fairly simple questions:

• Purpose: what outcome are we trying to create for the customer?
• Principles: how should we choose when several options could work?
• Priorities: what takes precedence when customer, commercial and operational interests conflict?
• Permissions: what can the agent decide on its own, and when must it ask the customer or a colleague?
• Red lines: what will we refuse to do, even if it would improve conversion?

The difficult part is agreeing on the answers and making sure they show up in the product. That takes brand, business and technology people working through real situations together. It also takes checking what the agent actually does, including the moments when there is no obvious right answer.

The brand book isn't going to disappear. We'll still need visual identity, voice and design. But an agent that can adapt to each customer needs something more useful underneath: a clear basis for making choices.

The more autonomous AI becomes, the stronger your brand needs to be.