As customer expectations rise and artificial intelligence (AI) capabilities mature, insurers have an opportunity to rethink the claims process. Rather than focusing primarily on what happens after a loss, they can use AI to anticipate risk, help prevent claims and deliver more personalized customer experiences.

In this episode of From Transactions to Trust, Thomas Rauschen, global industry lead for insurance, speaks with Guillaume Brincin, an expert in CGI’s AI and technology practice in Canada, about how AI is reshaping property and casualty (P&C) claims. They explore the shift toward agentic AI, the growing value of insurance data, opportunities for proactive claims prevention and the importance of reimagining the claims journey end to end.

Key takeaways

  • How can insurers transform claims from a cost center into a value driver?
  • How can AI help insurers move from reactive claims handling to proactive prevention?
  • Why should insurers reimagine the end-to-end claims journey rather than simply automate existing processes?
  • How can insurers bring the benefits of AI directly to customers?
  • What foundations are needed to scale AI across claims?

How can insurers transform claims from a cost center into a value driver?

Traditional claims processes have largely focused on what happens after a loss: receiving the first notification, assessing the claim and ultimately making a payout. AI can help insurers intervene earlier in the claims journey.

According to Thomas, insurers should think beyond processing claims more efficiently and consider how technology and data can help prevent losses in the first place.

“Claims are no longer about processing and payout; they must become a strategic lever for customer experience, risk prevention, and profitability.” — Thomas Rauschen

Agentic AI could accelerate this shift. Rather than simply using AI tools to support individual tasks, insurers can use AI agents to orchestrate workflows, with people providing appropriate oversight.

For customers, the potential benefits include greater speed, personalization and transparency throughout the claims experience.

How can AI help insurers move from reactive claims handling to proactive prevention?

AI’s ability to analyze large volumes of information can help insurers anticipate customer needs and risks.

Insurers have accumulated significant amounts of data over many years. When combined with information from sources such as telematics, photos and geolocation, AI can help turn that information into actionable insights, potentially identifying risks and enabling intervention before a loss occurs.

“The real value of AI is its ability to stay ahead of customers’ needs and deliver value at the right moment, ultimately strengthening satisfaction and trust.” — Guillaume Brincin

This means AI transformation cannot be separated from data strategy. The quality, availability and appropriate use of data will influence how effectively insurers can predict risk, personalize interactions and improve claims outcomes.

Why should insurers reimagine the end-to-end claims journey rather than simply automate existing processes?

Many insurers have already applied AI to specific activities such as contact centers and fraud detection. Insurers can now consider how to apply AI across the claims life cycle rather than to isolated activities.

AI can support first notification of loss, claims assessment, proactive prevention, fraud detection and feedback loops. Rather than adding AI to individual steps in an existing process, insurers can use it to rethink how those processes work together.

Thomas describes the distinction clearly:

“It’s about not putting patches on the current process but really reimagining your claims process with AI.” — Thomas Rauschen

Fraud detection illustrates both the opportunity and the challenge. AI gives insurers more sophisticated tools to identify suspicious activity, but bad actors also have access to rapidly advancing technology. Insurers can draw on their data and accumulated knowledge of fraud patterns to strengthen AI-enabled detection.

How can insurers bring the benefits of AI directly to customers?

Much of insurers’ early AI investment has focused on internal efficiency, where risks can be more easily controlled. As AI capabilities mature, Guillaume argues that insurers need to extend those benefits to customer-facing processes.

“We really need to go further and bring the AI gains to the client-facing side.” — Guillaume Brincin

That requires viewing claims transformation through the customer’s experience, not simply through operational metrics. Faster processing and lower costs matter, but so do transparency, personalization, proactive communication and trust.

The goal is not simply a more efficient version of today’s claims journey. It is an experience designed around what customers need before, during and after a potential loss.

What foundations are needed to scale AI across claims?

For insurance leaders considering where to begin, strong data and AI governance are fundamental.

Sensitive and personally identifiable information (PII) requires appropriate safeguards, while different AI workloads may call for different environments. Guillaume points to hybrid approaches that combine sovereign or on-premises AI with cloud-based capabilities, supported by clear data classification and governance.

Data quality is equally important. AI can amplify what already exists within an organization, meaning poor-quality data can quickly undermine AI performance.

With those foundations in place, insurers can experiment more quickly, extend AI across multiple parts of the claims life cycle and gradually bring AI into customer-facing interactions.

How insurers apply AI across claims could become a source of differentiation as organizations seek to deliver faster, more proactive and more trusted customer experiences.

Listen to the full episode to learn how insurers can move beyond incremental claims automation and use AI, data and end-to-end transformation to create more proactive, customer-focused claims experiences.

AI and the future of claims: From efficiency to customer value

Chapter 1: How can insurers transform claims from a cost center into a value driver?

Thomas Rauschen:

In today's episode, we will explore how AI is transforming claims in P&C insurance. To shine a light on this question, I'm thrilled to be joined by Guillaume Brincin, who is a leader in our AI and technology practice in Canada. Hi Guillaume, I'm delighted to have you on my podcast.

Guillaume Brincin:

Hi Thomas.

Thomas Rauschen:

Before we deep dive into the topic, let me start with an opening statement. For decades, claims have been the moment of truth in insurance, and as you know, even today it remains largely reactive and focused on payout after the loss rather than a prevention. Also, based on our own experience and discussions with clients, the first notification of loss and claims handling are still fragmented and slow, creating friction with customers and inefficiencies for insurers. But now, in a world of rising customer expectations, rapid technology advancements, and increased competition, I do believe the traditional model is no longer good enough.

Before we dive into the AI discussion, what is your perspective on the current state and how AI can help to overcome some of the customer frustration that we see across the market?

Guillaume Brincin:
The chatbot era is over. Since 2025, agentic AI is a keyword. And now, in 2026, it is basically what everyone should do with orchestrated agents, specifically to manage entire workflows. We are changing our paradigm from humans using AI to AI agents completing tasks with human oversight. This is basically the AI part, and for claims, we can see higher volumes.

Customers have high expectations for speed and efficiency, and this is where AI can help us by delivering hyper-personalization, hyper-customization, and greater transparency, especially during the global orchestration of claims. We can have prediction and more reactivity as well. And some of our clients are bringing the next best interaction, which is associated with the use of AI in claims.

Chapter 2: How can AI help insurers move from reactive claims handling to proactive prevention?

Thomas Rauschen:

So basically, what you were saying is that the old model is really more focused on payouts, right? Even with the old model, customers still expected speed, transparency, and quick payouts for their claims. Obviously, that changed. The expectations are even higher because we have AI; we can use automation. But equally, what we also see in the market, Guillaume, is that companies are thinking more and more about how we can help customers, how we can use our data, and how we can help prevent claims instead of only focusing on the payout. Do we experience the same?

Guillaume Brincin:

Yes, exactly. You said it: the word is "data" because we need to leverage AI's capacity to manage large volumes of data. It would help us predict and anticipate.  I talked earlier about next best interaction. The overall thing that AI can bring is that the process is always ahead of the customer's needs. And then value is delivered at the right moment, and it will increase satisfaction and trust. All this can be done by, as I said, managing a large volume of data, and that data comes from different sources; it can be IoT, it can be online behavior, it can be social networks as well. We are gathering a lot of information that only AI can help us manage and profit from.

Thomas Rauschen:

That's a good answer. Following the narrative that you just laid out, again, customer expectations are higher. But equally, we see this movement from just payer to prevention. Of course, modern technology and AI, you know, can help to automate your claims process and reduce the cycle times and reduce the operational costs, but beyond automation, AI is also able to unlock predictive and preventive capabilities, right? Can we talk about a few examples here?  When I speak to clients, I hear preventative ecosystems, data-driven risk anticipation, AI-enabled early intervention. Can you talk about a few examples here?

Guillaume Brincin:

It’s really a matter of the kind of data you have first; everything is based on your data quality and the level of information you can gather or aggregate with your client. So, it's directly linked to sensitive, non-sensitive Personally Identifiable Information (PII), of course. But when we are talking about prediction and especially claims prevention, it's really a matter of understanding what the clients are doing at the precise moment and anticipating what they will do next.

So, during a client's life, there are some major steps. You are getting your diplomas, then buying a home, buying a car, moving to another country, or taking a vacation. And regarding the overall advancement or progress in life of the clients, you can anticipate what kind of claims or predict what will happen, especially if you have that real-time information coming from auto capture, from photos, telematics, geo-location; you will have a very precise image of what the clients are doing and what kind of risks they are exposing themselves to. And you can take some preemptive actions to avoid specific claims and probably accidents also.

Chapter 3: Why should insurers reimagine the end-to-end claims journey rather than simply automate existing processes?

Thomas Rauschen:

That's a good answer because it sounds like, in order to have less friction, a digital and an AI-enabled customer journey, it's not only that companies need to look at AI. Of course, it's an important part, but equally, data quality and apps and new technology like auto-capture data, as you said, like photos, telematics, geolocation. So, what I understand you are saying is it's not only about AI; you also need to think about your end-to-end claims process and what kind of data and technology you need at certain points in the customer journey. Is that right?

Guillaume Brincin:

Yes, absolutely. We've been gathering information for years now, and we can see that most of this can't be managed or used properly, and the use of AI is clearly changing the game here because we can exploit or use that information, that data, quite efficiently, and then it can lead to actions, predictions and better overall services to our clients.

Thomas Rauschen:

That's a good segue into our next topic because I believe, as you just said, when we, for example, talk about telematics, right? Or photos, I think that has been around for the last, probably, like five to ten years, right? But now AI provides us with far more analytical capability in order to skim through the data and analyze the data and make the right decisions for the customer, but equally for the company. If we talk specifically about AI across the claims lifecycle, where specifically does AI change the game? Of course, the claims process is long, but from your experience, either in first notification of loss or in claims assessment, how specifically can AI help?

Guillaume Brincin:

Yes, for quite some time, clients and companies were focused on some parts of the claims lifecycle, especially for call centers, fraud detection or other specific parts of the lifecycle. But what we can see is that this is applicable to any part of a company where AI can be used, and the idea is that you need to have a holistic approach to AI throughout the whole claims lifecycle. Now AI is allowing us to create lots of POVs, POCs very efficiently in a few hours. You don't just have to focus on one part of the lifecycle. You can address it as a whole. So, you will have the first notice of loss, claim assessment, proactive prevention layer, claims inside of feedback loops; all those steps can be addressed all at once, and you will have massive gains at each of them.

If I can just bring up something you said earlier regarding data and the fact that we have more information and capability to address client needs.

One major area that is very interesting right now is fraud detection, because AI is bringing new tools to detect fraud and greater transparency into the process for clients. However, we need to keep in mind that criminals also have access to AI right now, so it's really a race to see which will be more advanced. But we have to keep in mind that insurance companies have a huge advantage regarding fraud detection, and it is once again their data and their overall history of fraud prevention, which will be very useful in the use of AI because it will bring all this knowledge, and basically, this is why we'll keep our insurance company ahead of that criminal behavior regarding fraud.

Chapter 4: How can insurers bring the benefits of AI directly to customers?

Thomas Rauschen:

That's a great example, and I think in our fourth podcast, Guillaume, we had exactly the discussion about how to build proper governance around AI. To prevent what you just described. But when I think about what you just said, of course, AI can be bolted on at a specific step in a claims process, right? But that basically means you still have your old claims process, which is a little faster and a little less frictionless. But the reality is what you just described: instead of looking at one step in the process chain, you really look at the whole claims process holistically and reimagine the whole process, and that leads you to faster resolution, again, almost no friction, but more importantly, increased customer satisfaction.

One thing I wanted to mention here, though, and I don't know if it's your experience as well. I also sometimes see that many companies discuss claims from an internal perspective. You know, we digitalized our claims process; we are more cost-efficient, et cetera. What is often missing, though, is that we need to put the customer in the middle. You really reimagine your claims process for the customer, not just from an internal perspective. Is that also your experience?

Guillaume Brincin:

Basically, the reason why companies and insurance companies specifically are focused on more internal optimization, or the way they are calculating how AI is helping us, is that once you are on the internal side, you are managing your risk; you mitigate risk more easily than when it's client-facing. Once you are facing the client, you can't predict what is going to happen. And if something goes wrong, it will obviously severely impact the whole company. So, this has been the global trend regarding AI and the insurance field. Most of the work has been done on the internal side. And what should be done right now, because AI tools are now gaining sufficient maturity to do so, is to bring those gains to the clients themselves. So, we can see some adjustments right now regarding the interface or some processes, but we really need to go further and bring the AI gains to the client-facing side.

Chapter 5: What foundations are needed to scale AI across claims?

Thomas Rauschen:

That’s a great answer, and I think we are aligned with that.  Before we wrap up with our takeaways from the podcast, when it comes to transforming the claims process and improving the overall customer experience, in your client conversations, what are the practical steps leaders should take tomorrow? Where would you advise your clients to start?

Guillaume Brincin:

The first thing that we are discussing with our client is that when we are addressing the claim lifecycle, there is obviously some concern regarding sensitive PII, non-sensitive PII and, globally, the overall data management through AI. So, some of our clients can be stuck at this point, asking themselves, "What should I do?" Can I put PII or sensitive PII in an AI model? The answer is no. But the first takeaway is that you shouldn't have only one provider, meaning only one way of doing AI in your company. It is best practice for strategic purposes and for the stability of AI in your company. 

So, you should have more of a hybrid approach with sovereign or on-prem AI use and then cloud-based AI, which is usually what companies are doing. And you should have some data classification regarding this, so you can be sure that your data is used safely with AI. So, this is one part, and once you have this kind of approach, everything is unlocked because you can do AI at any part of your claims lifecycle. And as I mentioned earlier, AI can help us now create proof of concepts in just a few hours. So, we just don't have to focus on quick wins now. We can broaden our view and use AI across different parts of the lifecycle. And as we just said, we should use AI on the client-facing side by taking small steps, but doing them faster and getting bigger very quickly, because AI now has the maturity to do so.

If you want to start, you will obviously have to adopt an AI launchpad or a way to get on the AI train. The first step is, as I said, to manage your way of doing AI. Be sure that you have good AI data governance and that the level of quality of your data is high enough for you to do AI, because, as everyone is saying, AI is just increasing the faults. AI is just accelerating or increasing what is already happening in your company. So, if your data quality is not as high as you would like it to be, then your AI model won't be performing as well as you want it to. So be very focused on that, and then you'll see you'll gain very, very rapidly in using AI.

Thomas Rauschen:

Perfect, and that's a good finish for our podcast. I would like to share my three key takeaways with you, and you can tell me afterward whether they resonate with you. So, my first takeaway is that claims must evolve from a cost center to a value driver. So basically, claims are no longer about processing and payout; they must become a strategic lever for customer experience, risk prevention, and profitability.

The second one is that AI enables a shift from reactive to proactive claims processing. So basically, AI transforms claims by enabling real data intake and smarter decisions.

 And the third one, for me personally, the most important one: success requires more than technology; it demands transformation. So, to unlock AI's full potential, insurance must rethink the processes end-to-end and obviously invest in data quality.

So, it's about not putting patches on the current process, but really re-reimagining your claims process with AI. Does that resonate with you?

Guillaume Brincin:

Yeah, I'm completely aligned with those three takeaways, and it should be kept in mind because this is where the competition will be in the next few months. This is where insurance companies will differentiate themselves from other companies and the competition globally.

Thomas Rauschen:

Thank you so much, Guillaume. I really enjoyed the session, and I'm grateful that you could make it today and share your insights.

Guillaume Brincin:

Thank you for the invite; I was very glad to talk with you about AI. You know, this is a subject that I particularly like.

Thomas Rauschen:

Thank you so much!

Explore more From Transactions to Trust episodes