CXO says that AI is changing not only the experience, but the customers themselves

Valto Loikkanen is the co-founder and Chief Experience Officer at Prifina, a company that builds personal AI tools.

We sat down with Valto to get a sense of how AI is changing CX. He said that the way customers engage with products is changing, and companies can no longer control the journey.

A focus on end-to-end experience

I’m Valto Loikkanen, cofounder and Chief Experience Officer at Prifina. My work intersects customer experience, product architecture, personal AI, and user-owned data.

At Prifina, we build personal AI products around a simple idea: AI should not only be what companies use to understand, target, or automate people. It should also empower individuals to control their information, relationships, decisions, and digital life.

This includes our personal AI twin technology, which allows individuals to create AI representations of themselves for public, professional, or audience-facing interaction. It also includes Souvéa, our personal AI companion app, which helps people understand what matters in their everyday life, such as messages, plans, tasks, personal context, and items needing attention. The experience is calm and trust-based: your data stays yours, you decide what Souvéa can use, and nothing happens without you.

Our users and customers include professionals, creators, experts, organizations, ecosystem partners, and everyday individuals who want AI to enhance communication, clarity, insight, and more meaningful interaction. Channels include web, mobile, public AI twin pages, onboarding flows, personal data connections, knowledge management, analytics, and partner-facing tools.

In my role, I focus especially on the end-to-end experience: how people understand these new AI concepts, how they get started, how they stay in control, how we build trust, and how AI can support better relationships between individuals, companies, services, and communities.

How AI makes customer experience more proactive

We moved customer experience from a mostly static journey to an AI-supported, conversational experience where users start from their own situations.

Traditionally, onboarding and product education are designed around fixed flows: Explain the product, show the steps, and guide the user toward predefined actions. With AI, we made that experience more adaptive. Instead of assuming every user has the same question, the experience can respond to what the person is trying to understand or achieve.

So in our products, we focus more on conversational entry points. Users do not need to fully understand the product category first. They can ask questions, explore examples, clarify how their data is used, and gradually understand what the product can do for them. That means users might start with a simple question like “What needs my attention today?” rather than navigating through a traditional dashboard.

That sounds simple, but it changes a lot. The product experience's role changed. It is no longer just a set of screens and instructions. It becomes more like a guided relationship, where the product can explain itself, respond to context, reduce cognitive load, and help the user feel more in control.

This improves clarity, lowers the barrier to first use, and helps us see user intent much earlier.

It's also a major shift for customer experience: AI does not just automate support after something goes wrong. It becomes part of building trust, clarity, and confidence from the beginning. And it also provides better signals about what people care about — instead of only tracking clicks or drop-offs, we can understand the questions behind the behavior.

An end-to-end AI-powered CX workflow

Here's the end-to-end AI-powered CX workflow that loops user interaction, insight generation, and experience improvement.

1. The user interacts with one of our AI products using natural language.
2. We interpret interactions. AI helps us understand individual interactions and the patterns behind them. Are users asking what the product knows about them? Are they unsure how to get started? Are they trying to use the AI for a use case we did not clearly explain? Are they asking for more control, more examples, or more reassurance?
3. From these signals, we generate CX insights. These insights can inform onboarding, product language, permission flows, help content, feature prioritization, and how the AI responds. For example, if repeated questions show users are unsure about data use, the solution is not just a support answer. It may mean changing onboarding, adding clearer permission explanations, or adjusting how the AI introduces itself.
4. Finally, we close the loop. We update the product experience, then observe whether the same confusion decreases or new questions appear. In that sense, AI not only powers the customer-facing interaction but also helps us continuously understand where the experience creates clarity and where it creates friction.

Why companies no longer control customer journeys

As a result, I realized that the company is no longer the main designer of the customer journey.

For a long time, companies have designed customer experience from their side. We map the journey, define the touchpoints, create the funnel, write the support flows, and decide how the customer should move through the experience. Even when we talk about customer-centricity, companies usually still own the structure.

AI changes that. When people can ask questions in their own words, use personal assistants, compare options, summarize information, and eventually have AI agents act on their behalf, the journey becomes much less linear and much less controlled by the company.

That means CX can no longer only guide people through a designed path. It also has to support people who arrive with their own context, their own AI tools, and their own way of understanding the relationship.

Where AI excels and where human judgment is required

Overall, we use AI especially where it identifies patterns, reduces complexity, and makes the experience more responsive for the individual.

But humans still need to provide judgment around meaning, trust, ethics, and direction. Humans still decide what kind of relationship we want to create with users, what should never be automated, what data use is acceptable, and how much agency the AI should have. Journey design also remains strongly human-led because it is not only about efficiency. It is about emotion, confidence, expectations, and the overall promise to the customer.

AI should inform and support more CX decisions, but it should not quietly take over responsibility for them. Especially in personal AI, the most important design question is not “What can we automate?” It is “What should remain under the individual’s control?”

How AI raises the stakes in CX

AI does present some challenges. For example, it can raise expectations too quickly. When a product feels conversational and intelligent, users may assume it can do more than it currently can or understand more context than it actually does. This creates a higher responsibility to communicate boundaries clearly.

Another challenge is trust. In CX, a wrong button label is annoying. A wrong AI answer can feel personal, misleading, or unsafe, especially when the product deals with people’s own information. That raises the quality bar. Design for reassurance, transparency, permission, and human control from the beginning.

So far, the biggest lesson is that AI can significantly improve CX, but only when it creates clarity and confidence. If added mainly as automation, without a clear trust model, it can also create confusion faster than traditional software.

Why proactive AI assistance should be used with caution

Be cautious when using proactive assistance with AI.

It is tempting to think users always want AI to notice things and suggest actions. In reality, proactive help only works when it is very well calibrated. Too many suggestions feel noisy; too few feel useless. If the suggestion is wrong, it can quickly reduce trust.

As a result, we now think more carefully about calmness, timing, permission, and user control.

How to redesign onboarding for AI experiences

With all this in mind, I would redesign onboarding around a few practical strategies.

The first is conversational discovery. Instead of forcing every user through the same onboarding path, the product should ask simple questions and adapt to the user’s situation. I covered this above.

The second is progressive permission. Users should not connect everything at once. They should start small, see value, and then decide whether to add more context or data. This is especially important in personal AI, where trust grows step by step.

The third is example-led onboarding. AI can feel abstract until people see practical situations.

The fourth is an explicit trust layer. The onboarding should clearly explain what the AI knows, what it does not know, where information comes from, and what remains under the user’s control. This should not be hidden in a privacy policy. It should be part of the experience. And make escalation paths clear as well.

The fifth is onboarding analytics based on intent, not only clicks.

Why CX leaders should start with narrow AI use cases

I wish I had known how important it is to start narrower.

With AI, people quickly imagine broad, almost unlimited use cases. That can create excitement, but it can also create unclear expectations. A focused pilot with a very specific promise is often more valuable than a broad pilot trying to show everything AI might eventually do.

My advice is to define the experience promise before the technical scope. What should the user feel after using it? More clarity? More confidence? Less stress? Better control? If that is clear, evaluating the AI implementation becomes much easier.

Valto Shares

My advice is to define the experience promise before the technical scope. What should the user feel after using it? More clarity? More confidence? Less stress? Better control? If that is clear, evaluating the AI implementation becomes much easier.

How AI changes CX teams

AI has made the CX function more cross-functional, more product-embedded, and more agent-assisted.

In a traditional setup, customer experience can easily become a separate layer: research, support, feedback, journey mapping, and service improvement. With AI, that separation does not work as well because the experience itself is partly shaped by the intelligence inside the product.

That means CX now needs to sit much closer to product design, data architecture, AI behavior, privacy, onboarding, and support.

It has also changed what we consider a team. The CX team no longer consists only of people. AI agents can now act as team members that help with research, synthesis, pattern detection, content drafting, onboarding review, support analysis, and product feedback loops. They do not replace human judgment, but they expand the team’s capacity and help us work with more context than a small team could manually process.

For us, this has made the CX role more strategic and architectural. It is less about optimizing isolated touchpoints and more about designing the relationship between the person, their data, the AI, and the product.

The skills that matter have also expanded. We still need strong product thinking, UX, communication, and customer understanding, but we also need better judgment around trust, permissions, data use, AI boundaries, agent behavior, and the emotional impact of automation.

How customers are changing and what CX leaders need to do about it

My advice to CX leaders is to stop thinking about AI only as a tool for automation and start thinking about it as a new relationship layer between people, companies, products, and services.

Many organizations will use AI to reduce cost, speed up support, personalize marketing, or automate service processes. Those are useful, but they are not the whole opportunity. The deeper change is that AI will reshape how people understand choices, make decisions, interact with brands, and manage their own lives.

That means CX leaders need to ask bigger questions. Are we using AI to make the company more efficient, or are we also making the customer more capable? Are we personalizing experiences in a way that builds trust, or just optimizing conversion? Are we giving people more control, or are we making the system more invisible and harder to understand?

Winning CX organizations will combine intelligence with trust. They will use AI to reduce friction, but also to create clarity. They will personalize, but with permission. They will automate, but not hide responsibility. They will design experiences where customers feel more informed, more respected, and more in control.

This is especially important because customers will also have their own AI tools. In the future, the customer side of the relationship will become more intelligent, too. CX leaders should prepare for a world where experiences are not only designed for people, but also for the AI agents and personal assistants that act on their behalf.

Follow along

You can follow Valto Loikkanen's work on LinkedIn. And check out Prifina.

More expert interviews to come on The CX Lead!