Cebu Pacific Air

Glenn Ong

Customer Trust Steward

Everybody talks about AI. Far fewer people talk about knowledge management.

That is where many customer experience transformations begin to drift. Companies are moving quickly to deploy AI, especially in service channels where speed, scale and cost matter. The promise is clear. AI can answer faster. It can serve customers at any hour. It can reduce pressure on frontline teams and make support more accessible.

But AI is not magic.

AI reflects the quality of an organization’s knowledge ecosystem. When that ecosystem is clear, governed and consistently updated, AI can become a powerful layer in customer experience. When it is fragmented, outdated or inconsistent, AI can deliver the wrong answer at scale, and do so with confidence.

The risk is not theoretical. In customer experience, inconsistency is one of the biggest trust breakers. Whether a customer reaches out through a chatbot, an agent, social media or a physical touchpoint, the answer should be consistent, accurate and grounded in the same source of truth. When information lives in too many places and no one fully owns it, that consistency becomes hard to protect.

Before the Bot, Fix the Source of Truth

Many organizations underestimate how fragmented their knowledge is until they attempt AI deployment.

Policies may sit across multiple teams. Operational updates may move faster than communication workflows. Commercial, operations, digital and customer service teams may each hold part of the answer. Frontline teams may know what works in practice, while customer-facing channels still reflect older language. From the inside, the problem may not look serious because people have learned how to work around it.

AI removes that comfort.

Once automation enters the picture, every unclear policy, outdated answer and conflicting interpretation becomes more visible. Worse, it becomes scalable. If the knowledge base is incomplete, AI does not quietly fill the gap with organizational wisdom. It may surface the same confusion with confidence.

That is why knowledge management must come before AI.

The Hard Work is Clarity, Governance and Ownership

Before automation, organizations must solve for clarity, governance, ownership and consistency of information.

These are not glamorous words, but they decide whether AI can be trusted. Is the policy accurate? Is it written in a way customers can understand? Who owns the source of truth? Who approves changes? How fast can information be updated during disruptions? Are frontline teams aligned with what AI is saying? What happens when AI gets something wrong?

AI systems rely on structured, current and governed information. If policies are unclear, outdated, conflicting or incomplete, AI will surface those problems. Poor knowledge governance increases hallucination risks and weakens customer trust.

Knowledge management, therefore, cannot sit only within support functions. It requires cross-functional ownership across operations, digital, customer service, marketing and commercial teams. Customer experience is not shaped by one department alone. The knowledge behind it cannot be owned by one department alone, either.

It is also not a one-time clean-up before launch. Knowledge governance requires continuous ownership, maintenance, review and improvement as business conditions, policies, routes, disruption patterns and customer needs evolve. The source of truth has to move with the business, or the customer experience will quickly fall behind it.

The need becomes sharper in aviation, where policies, operational conditions and disruptions change quickly.

A customer asking about rebooking, refunds, baggage, checkin, travel documents or schedule changes is often asking under pressure. Their plans may have already been disrupted. They may be worried about missing a flight, losing money or not knowing what to do next. In those moments, they are not looking for a clever answer. They are looking for certainty.

If AI gives an answer that sounds confident but does not match the actual policy or operational reality, the damage is not limited to one interaction. It affects trust in the airline, the channel and the entire service experience.

For this reason, knowledge management is not just documentation work. In reality, it is operational discipline and customer trust management.

At Cebu Pacific, our journey toward generative AI did not start with the chatbot itself. It started with strengthening the airline’s knowledge foundation.

As we worked toward launching one of the first generative AIpowered customer chatbots among low-cost carriers in Southeast Asia, we had to review what the system would rely on: our policies, customer-facing information, common concerns and operational realities.

We reviewed and simplified policies. We standardized customerfacing information. We improved knowledge governance. We also focused on the most common customer concerns, making sure the answers were operationally accurate and easy to understand.

The work may sound basic. It was not. It required teams to align around what should be said, how it should be said and how quickly it should change when the operating environment changes.

AI Was Never Meant to Replace People

AI was positioned not as a replacement for humans, but as a way to improve speed, accessibility and consistency of support.

This distinction matters. In customer experience, there will always be cases that need human judgment. Some situations are emotionally sensitive, high-risk or operationally complex. A passenger dealing with a disruption, a personal emergency or a complicated service concern may need more than an automated answer.

Human escalation paths remain important. AI should help customers get reliable answers faster, while allowing people to focus on situations that need empathy, discretion and context.

The goal is not to remove the human element from customer experience. The goal is to make the experience more dependable, especially when customers need clarity quickly.

AI Maturity is Really Trust Maturity

One of the biggest lessons from this journey is that AI implementation is not purely a technology initiative. It is also a customer experience, operations, communications and change management initiative.

True AI maturity is not measured by how advanced the tool sounds. It is measured by how trustworthy and operationally reliable the experience becomes for customers

A fast answer is useful only if it is accurate. A digital channel is valuable only if it is aligned with the business. A chatbot is impressive only if it can help customers without creating confusion.

The real challenge is not simply deploying AI. The challenge is operationalizing trust at scale.

Wisdom from Experience

Before deploying AI, leaders should ask practical questions. Is our knowledge accurate? Who owns the source of truth? How quickly can we update information during disruptions? Are our frontline teams aligned with what AI is saying? What is our process when AI gives a wrong answer? How do we keep improving the system after launch?

These questions may sound less exciting than technology demos, but they determine whether AI succeeds in the real world.

In the coming years, knowledge governance is not a onetime cleanup before launch. It requires continuous ownership, maintenance, review and improvement as business conditions, policies, operations and customer needs evolve.