The marketing AI paradox

Marketing is experiencing profound change thanks to AI, and at a speed faster than most expected.

A 2025 Gartner survey found that 65% of CMOs expect AI to dramatically change their role within the next two years. That’s two thirds of CMOs expecting not just the day-to-day activity but the role itself to transform.

This means that marketing leaders without a high level of AI marketing literacy will be at risk. AI is reshaping how organisations grow and compete.

The paradox is that marketing’s use of AI is already widespread. Marketing teams today are using generative tools to produce content and automate campaign workflows. Coca-Cola is a notable example (with some public backlash) of a business unashamedly using generative AI in campaign visuals and ads.

However, a lot of today’s usage is largely tactical and focused on short-term efficiency gains. Those gains are more operational than transformational. They have value but are unlikely to change the trajectory of marketing in a way that executive peers might expect. This increases the pressure on marketing leaders even as they embrace AI.

How do marketing leaders use AI to drive long-term commercial advantage? How can the promise of AI innovation, such as reshaping customer relationships and creating new experiences and growth, be realised?

To answer this we must first look at what AI is already doing to customer relationships.

AI is already affecting the customer experience

Automating processes is only one consequence of AI’s arrival in marketing. The more significant shift is already underway, where AI is increasingly appearing inside the customer experience itself.

Recommendation engines shape product discovery and automated systems personalise messaging and offers. Conversational interfaces increasingly sit at the front of support and sales interactions. Agentic AI is starting to mediate how customers interact with organisations.

Of course, all this use of AI still requires a great User Experience for engagement and adoption, but the trajectory is clear. AI is quietly reshaping the relationship between brands and their customers, often without marketing leaders deliberately designing how that experience should look and feel.

When the AI integration works well, the experience improves. Marks & Spencer, for example, has deployed AI-driven tools across online journeys to improve product discovery and personalisation. The goal is for customers to find relevant products more quickly, and the interaction to feel like a tailored experience rather than robotic.

However, poor AI integration can have devastating effects. Consider a website chatbot that triages customer questions before passing them, if needed, to a sales or support person. In the physical world, few executives would place a robot at the entrance of their flagship retail store and require every visitor to interact with it before speaking to a human. The risk would be unimaginable. What if it said something untoward, or angered a customer? Yet a website chatbot that screens customers has the same influence and affects customer relationships at scale.

Negative AI effects are newsworthy effects. As one example, Air Canada’s chatbot provided a passenger with incorrect refund information, which the airline later refused to honour. The case ultimately reached court (and mainstream news), where the airline was held responsible for the actions of its own AI system. Another example in 2025-26 was xAI’s Grok being able to generate non-consensual sexualised images, attracting lawsuits, regulatory scrutiny and broad public backlash.

Customers do not want a robotic, unsettling AI experience. They want an experience worthy of the organisation and its values.

How do marketing leaders shape the AI experience so its impact is positive? How do marketing leaders avoid outsourcing their brand to AI?

Introducing Human-Centred AI

Human-Centred AI (HCAI) is the discipline of designing AI systems that are responsible, reliable, and trustworthy. It shifts the conversation from technical capability to commercial and experiential outcomes.

At Adjacent Possible, we see HCAI as the operating model for modern marketing, and the only way to transform customer experiences in a positive and commercially-sound way.

Many organisations begin with AI from a technical standpoint. Data teams experiment with models, technology teams deploy tools, and marketing inherits the outcomes.

What is needed is for marketing leaders to feel empowered to influence or dictate the outcomes of AI initiatives.

HCAI directly supports core marketing objectives:

• Customer trust and brand perception
• Customer experience quality
• Conversion and engagement
• Retention and lifetime value

It achieves this by evaluating AI through customer and brand outcomes.

HCAI asks questions like:

• How does the proposed AI solution shape behaviour?
• Does it build trust, or erode it?
• How reliable and consistent is the experience?
• What are the harms and risks of this solution?
• In what ways will it change how customers experience our brand?
• Have we managed its limitations so we increase our chance of better ROI?
• How will it directly influence KPIs and strategic goals?

Examples of an HCAI approach are already commercialised. For example, Spotify’s recommendation system embodies HCAI principles because it does more than automate the discovery of new music. It creates a sense of personal relevance and emotional connection, helping users feel understood by the platform.

As another example, in our collaboration with Envolve Tech, the challenge was not simply implementing AI tools to create more natural-sounding chatbot responses. It was designing the interaction between humans and the technology so that people could understand, trust, and confidently engage with the system. By focusing on experience design, clarity, and human interaction, we helped ensure that AI enhanced rather than disrupted the virtual shopping assistant experience.

This is the core principle of Human-Centred AI: AI becomes valuable when it is designed for people, not just for efficiency and performance.

How marketing leaders can benefit

For many organisations, we find the first step is two-fold:

• Build the marketing leader’s AI literacy
• Make sense of the AI that is already encroaching on the organisation.

For the latter, most organisations are in one or more of these situations:

• A large vendor has promised generic integration of AI throughout the business (e.g. Microsoft Copilot for every employee)
• Several technology-led AI exploratory projects are in flight, with many not making it to production
• Multiple individuals or teams are using ad-hoc, off-the-shelf AI solutions (such as ChatGPT) to improve their workflows

Without applying a strategic lens, these initiatives quickly become fragmented and underused, with poor ROI.

Instead, to move forward, marketing leaders need three things:

1. AI literacy. Understanding what AI can and can’t do, and where it creates experience and brand risk, allows marketing leaders to participate meaningfully in strategic decisions rather than being secondary to technology. Required skills include understanding different AI types, knowing AI’s capabilities and limitations, knowing which customer journey parts benefit most from AI, and spotting brand and reputation risks from proposed AI projects. We provide external consulting that builds AI literacy and skillsets, and reviews current AI projects and strategy with a HCAI and marketing lens.
2. Reframing of AI around marketing outcomes. In-progress and upcoming AI projects should not be defined by tools or models. They should be defined by the experiences and commercial results they are designed to deliver. We workshop this on a per-programme or per-project basis using an HCAI approach, asking the unasked questions and framing AI initiatives in terms of brand impact and customer experience.
3. Alignment across product, technology, and marketing teams. AI transformation is inherently cross-functional. Marketing leaders must manoeuvre to shape or even direct this transformation. This is best achieved by using the above two improvements (AI literacy, reframing AI projects) to give marketing leaders more natural influence in dictating the direction of AI projects and getting the right type of reporting on their progress. Marketing leaders get a seat at the strategic AI table when they are consistently and expertly translating AI into commercial returns.

Your organisation doesn’t need to be first at AI, but it does need to do it right. At Adjacent Possible, we support marketing leaders making the transition to becoming the strongest voices for commercially-focused, risk-managed AI that delivers a great customer experience.

If you are navigating these questions, we welcome a conversation. Contact us at [email protected] to discuss your unique situation.