EP 30 - Why AI Needs Human Governance and Business Discipline to Matter in Insurance
AI is everywhere. It has moved from technical circles into daily conversation. Open a newspaper, walk into any boardroom, or simply unlock your phone and you will be confronted with predictions about automation, productivity, and the imminent transformation of entire sectors. Insurance is no exception.
In this Episode:
AI in insurance creates value only with human governance and business discipline. François explains why insurers must align technology, people, and compliance to achieve real efficiency and long-term impact.
I have followed AI since my engineering studies. I have always been fascinated by its potential, not as a spectacle but as a tool that can work with people rather than around them.
What interests me is not the idea of replacing human capability. What interests me is creating the conditions for progress. In insurance, progress comes from the interplay between people, process, and protection.
What we are witnessing today is the first time AI has reached broad societal awareness. A billion people touch AI tools in one form or another. This scale creates both excitement and pressure. It invites the misconception that AI will handle everything while humans step aside. But the reality inside insurance companies tells a different story.
Insurance is an industry built on responsibility. It is regulated for a reason. It protects individuals, families, companies, and societies at moments of vulnerability. Because of this responsibility, no insurer can operate without human governance. Compliance, oversight, risk monitoring, ethical judgment, and client trust require human accountability. Without that foundation, the system loses its integrity.
This is why AI cannot simply be deployed in isolation. It must be guided by a clear business case, by value for money, and by disciplined leadership. This is especially true for the senior executives who sit at the core of the industry and carry the mandate for change. These leaders are analytical, intentional, and deeply aware of the consequences of technology decisions. They fear stagnation and misalignment far more than the noise of a new trend. They want progress that lasts, not progress that pleases.
Inside insurers, AI can serve many functions. It can improve the client experience. It can support operations, accelerate claims, strengthen the use of data, and create more consistent decision-making. But none of this happens automatically. It requires teams who understand how to use the tools, leaders who select the right projects, and governance that ensures compliance at every step.
The real work is not the coding. The real work is the integration.
When AI is introduced inside a silo, it becomes an island. The people outside the silo resist it because they cannot see its relevance to their own responsibilities. This creates friction, fear, and waste. But when AI is introduced as part of a shared journey, with education and transparency, something very different happens. People begin to recognise that AI can carry the heavy lifting. It can remove repetitive tasks and allow them to focus on judgment, relationship building, and the human elements that matter most.
AI used well strengthens people. AI used poorly isolates them.
This is why governance is not a limitation. Governance is an enabler. It ensures that AI remains aligned with the mission of the organisation and the expectations of society. It creates the guardrails within which creativity and efficiency can flourish.
The insurers who will lead the next decade are not necessarily the ones with the largest budgets or the most complex models. They will be the ones who understand that AI is a team member, not an autonomous force. They will choose projects with discipline. They will confront the fears of their teams with empathy. They will build environments where machine intelligence and human judgment work in concert.
Transformation in insurance does not happen through technology alone. It happens when technology serves a coherent strategy, a responsible governance framework, and a human-centric culture.
Progress begins when humans and machines move forward together.
Timecode:
00:00 Introduction to AI
00:20 AI's Growing Presence
00:34 Personal Connection to AI
01:30 AI in Society
02:12 AI in Insurance
03:35 Challenges and Risks of AI
04:15 Optimizing AI in Insurance
05:36 Teamwork Between Humans and AI
07:07 Conclusion: Efficiency and Collaboration
François Links:
Apple Podcast
Transcript:
So, um, topic now is ai. It's not the first time that, uh, we, we, we speak about that, but it's a very important topic and, um. It's overwhelming now everywhere. So just open your phone or look at the media. Um, everyone speaks about ai. It's a bit like we spoke about ai, uh, internet, um, like 25 years ago or maybe a bit longer.
I can't remember exactly, but a while back, let's say. But, uh, much stronger. So it's, it's, it's overwhelming. I, I like ai. It's something that I, um, I, I studied. Uh, during my, my, um, engineering degree and, uh, which has been close to me, uh, one way or another for a while, sometimes it was, um, it was coming back from, uh, from, from, uh, ex.
Student, core student, uh, they were asking, you know, are you interested also when I was thinking about this, but, um, I don't think that immersing myself into the building of AI was, uh, was what I want. What I like very much in AI is the ability of, um, creating. Efficiency, creating, um, the, uh, the, the condition, um, to work for humans and, and, and machine, uh, to work together, which has been, you know, increasingly useful since, uh, JGPT was released and all the models and, and all the elements.
Of of AI have been made available to the public. Something that was, uh, previously confidential and pretty much like far away sometimes thinking, oh my God, I've seen a robot, or, you know, there's some improvement. Or look at that logistic firm. They have those little, um, trolleys that go from one place to another place.
But that was very much into the, uh, the domain of the expert and the specialist in one or several fields. But now it's become a, um, a society. Uh, feature where almost a billion people using JGPT today, um, on a regular basis is, is the scale is, is, is, is enormous and, um, therefore. This is overwhelming, overwhelmingly spoken about, and it's something that of course, uh, impacts, uh, the insurance company.
It impacts the insurance company on several level from the outside world. But in and inside the insurance company. So an insurance company could virtually be, uh, fully operated by, by ai as long as you have humans next to it that drive AI into doing what, what has to be done and, uh, is supported by the legislation in place.
So, uh, it's very clear that an insurer cannot just simply be, uh, fully AI based without having human intervention without. Having human control, governance and, uh, all the functions that are described in the law governing and, and securing the fact that the, the AI based insurer will be, uh, fully under compliant with the legislation and under control and, and, and therefore for stay in the long term.
But, um. Tradition. Insurance companies, of course, have a, you know, AI in their eye, uh, for, for a long time now, and that they've invested some that the biggest groups have established, you know, VC funds that are investing in, in AI companies or in the ecosystem of ai. And, uh, others are using simply AI to a certain degree.
Uh, some more, some less. There is a big danger, of course, in, in AI is that it can become very, uh, very, uh, time consuming, could be, uh, uh, money consuming. And, uh, there has to be a result at the end. And the hype today is of course the expect, you know, lays the basis, uh, that the expectations are. You know, AI will do everything and we will do nothing, but it's not exactly what happens.
Uh, there's a lot of. AI project that, uh, that are canceled, uh, that are readjusted simply because the outcome is not what was expected. AI is not there to do everything. AI is there to be a, a part of, uh, a team between humans and ai. Insurance companies are using AI for, um, better client experience. For, uh, better efficiency in terms of, uh, relationship with the, the client on the operations or in the back office, in the, in the claim management, uh, optimizing the data based, not the database, but the data based or fact-based, uh, decision making to accelerate the, uh, the outcome and, and, and, and create more consistency.
So the, there's a lot of element there and, and, and the clever insurers are the ones that, um. Select the right project to create the right outcome so that there's value for money. If there's no value for money, why would you do it? So there has to be value for money and there's trial and error. So laboratories are there to be, to, to, to be, um, um, used in, in, in, in of which the outcome is simply, you know, that works.
That doesn't work. Dangers and risks are here and there, but that has to be taken by. Competent teams and um, also the, the, the teams in place at, in insurance companies, they need to accept it. Uh, how do they accept that when they realize that they are, um, they can team with, with AI tools? That will accelerate AI based tools, et cetera, you know, that will accelerate and improve their outcome.
So the, the tools can create, uh, the condition of teamwork where the AI does the heavy lifting. In, in, in, in another podcast I spoke about the heavy lifting where, uh, team of individuals need to work together in, in a, in, in a, in the right way. So relying on several people to do the heavy lifting is easier to, to carry something within.
10 people in their shoulders than than one single person. So the same is, is with ai. And when, when it works, when the humans and the machine can work together, then the outcome is, is very positive. So the responsibility of the management of the, the, the people promoting AI or the employees promoting AI in in a company is as much.
As convincing and helping the other employees to use AI to understand how AI can work and how can they improve their outcome, break the fear, and give the right condition, as opposed to simply work on AI on their side in the, in the little, um, silo or large silo, depending on, on, on, on, on the wear, but.
Where silo simply creates a lack of interface. And lack of interface means that the people outside the silo will not accept it and will not be wor working with it. And that's a waste of money. That's a waste of effort, and that's a waste of time. So the message here is insurers are using ai, de efficiency is key.And, um, create a business case, create value for money. Add the insurers, make sure that humans. End machine are working together, not, uh, that uh, there are silos in fear building.