Transcript
service provider ecosystem. Recording live here at Mobile World Congress Barcelona 2026. From securing AI adoption to threat intelligence, business strategy, security operations and value-added services, we bring expert insight to help service providers grow, stay resilient and ahead of the curve. So let's dive in. I'm your host, Ronen Spier, Director of Technical Solutions and Marketing at Fortinet. And I am joined today by Marcus Hecke from Vodafone. So Marcus, great to have you with us. Thank you for joining. And I want to talk mostly about AI and you as you engage with your enterprise customer, Vodafone enterprise customers, what would be your assessment as to how do they adopt AI? They're really adopting AI in their businesses, in their workflows, or is it more of trying and playing with it? I think currently it's trying and playing with some very good initial, I would say cases already, delivering real business value. So I'm very often asked whether this AI topic is overrated or underrated. To be honest, I believe even if it sounds a little bit strange, it's both because it's very much overrated, I believe, on a very short term. And most likely it's completely underrated on the mid to long term. What I mean with that one is, if you look, there are real use cases delivering quite a significant value. Just to mention a few ones from the German market, for example. We have the Insura Allianz, and they already do fully automated AI driven claim settlement for smaller damages, really helping them a lot to drive efficiency. We see a very similar efficiency improvement with our car manufacturer, BMW. BMW is using AI for doing quality assurance for new cars, helping them a lot to drive efficiency. There is Zalando, it's the retailer being very much into clothing and doing the fitting and the proposing of new clothing for their retail customers. And I could continue, right? You have good examples in the financial industry, of course, in all the kind of service industries, and, and, and. So there's real business value. But if you look into the number of customers and the multitude of business processes, I think overall we are still more in the initial phase of AI. Are there specific sectors, businesses, business area that you see common eye value and return on investment for AI, or we cannot make that classification? Oh, I think what you definitely see is that there is a very different adaption rate across different industries. And just to give a few examples, the more you are in the knowledge work, right? For example, the consulting business, the media business, the telco business, everything what has to do with IT, especially software programming, is already significantly, I was adapting AI. You see other industries, which are more in the early phase, which has to do a little bit more around the manufacturing piece or energy and oil and gas. So this kind of utilities and so on, they also doing experimenting with AI. But I think the, if you look from the overall value creation, the portion which is driven already via AI is much lower than the examples I mentioned before. Okay. Okay. So even looking at these in overall, what are the key factors, if we can, if you can identify them that make their AI adoption a success or a failure? Yeah. I think there's a multitude of topics, which is very important. I think, and you hear that always, if you talk to people adapting AI, it's about the data. It's about the information. So you need to structure your data. You need to have cleaned it up because if you want to really use it and drive value from data, you should ensure that your data are meaningful. So I think this is across, I would say all the different kinds of industry. I think that is maybe the number one, but of course there are other aspects as well. Another one is of course the people, the people who need to be, I would say, educated and encouraged to really use that. Many people in different companies may be afraid of AI because they fear they may lose their job or they don't like the changes which are introduced. So you need to be very, I would say, be aware of that. It is not just a simple technology introduction, but it's a change management. People need to work differently. Business processes will change sometimes dramatically and so on. So you need to explain a lot. You need to have champions working. You also need to involve people even in the development of that AI, call it tools or capabilities, and you really need to consider that as a major change process. That is the second one. The third one I want to mention is AI is very often seen as you put AI and then AI is solving everything. But what you really should do is you should be very much aware of which business process, which problem you want to solve, and then you should work on to develop an AI solution which is exactly more or less looking into that specific problem and of course solving or helping to drive efficiency here. And very often it is more thrown into a company rather than having a very clear target or a goal being defined, what you really want to achieve. These are the three and maybe core topics I want to mention. And with regard to this, you mentioned processes awareness, which would bring me to the question of AI governance. Do you see enterprises develop and have in place AI governance as they adopt or even prior to them adopting AI and with this understanding, this is what I want AI to do, this is the process? And are they aware to the risks that might be associated with what they're trying to do with AI? Yeah, I think there may be even two very opposite, I would say, positions you can take. There are some companies which are so much aware of the governance and they are so much afraid that they restrict the usage of the AI so much that it only provides very limited value or which creates huge hurdles for implementation of AI. That is the one side. The other side is if you have less educated, they may completely underestimate the risks coming with AI, especially if you think about the new tech surface for cybersecurity or uploading of data into public cloud environments. I personally believe that especially on the employee level, very often the urgency or there is no real feeling that if they provide company insights, company data to public AI platforms, that this may be used to train the models or to be used also then for creating answers for others. And this, of course, may provide a huge risk for the company. So I think there are both sides. Personally, I believe there are many companies trying to get their hands around it. Most of them, I talked to, they are developing their set of government's rules on how to handle it. I think some of them more advanced than others. And some, as I already mentioned, may be even so strict that they may not get the business value which they could. So to close it off, and you mentioned in the beginning, you mentioned long term benefits. So what would you expect? What are the cool things and the impacts of long term AI that you expect and you want to see? Yeah, I think you already see the first indicators for what's upcoming. And this is very much the agentic development. You know that the model context protocol now provides the capability to more or less link agents to platforms, to even physical environments. And this will provide the next, I call it the next wave of usage of AI. So you can have completely automated agent driven execution of even rather complex topics. And if you develop an agent which can interact via this kind of protocols with other systems, then the next level, of course, will be agents orchestrating agents, right? So you can even link very different tasks or business processes to even more complex tasks and processes. There's two things I believe is very, very important for the decision making. I think we still should have the human in the loop, as it is called, right? But in principle, you already see that many of that even complex situations can be handled and solved via AI. And as another indicator, you may think about the open claw, which has recently got such attention because it shows how creative AI can be in just trying out to find solutions to a problem, in simply trying different ways to finding the best solution to a question you have asked the agent. I think we all also have seen that there is a huge risk if you don't give boundaries to this kind of approaches, because it may try things which you never want this agent to do. And of course, we already touched on that security topic and so on. So you need to develop this governance framework for agents. But I personally believe it gives you an idea what is possible if we will do the next steps and can define these boundaries for agents and have more strict governance so that you really can deploy that also in a more business like enterprise environment and not just trying it out on your personal PC, right? And last but not least, if you then think further and link that to the physical world, to the robotics and so on, I think there's a lot of fantasy what can be done. And as I want to recall one of my earlier statements to say it may be underestimated on the mid to long term, I believe you may see this kind of developments within the next year. And also to draw a conclusion to some of the other technologies we have seen, if you think about the mobile industry. Initially, we had a huge hope in all these different business models. Then you had more or less the depression because it didn't work as expected. But now all what we discussed many years ago is reality. The same happened with e-commerce, right? After the burst of the bubble, right? There was a lot of people very depressed about it. But now we have all the business models. Now, if you just continue along that line, think about AI. We have the initial ideas, big hopes and now we see it's not so easy to realize but what may be possible on the mid to long term. Exactly. And so I think there's a lot of hope and they will come true if we do a trustworthy AI and secure AI. Absolutely. And I only can encourage everybody, meaning especially, of course, businesses to really look into it and really to be open. Of course, be careful because of all that security and governance topics they need to consider. But they definitely should not miss out on that opportunity. Markus, thank you very much for your insights and for your time and for joining us. Thank you so much. Thank you. That's all for this episode of Fortinet On Air, recording live at Mobile World Congress Barcelona 2026. Fortinet On Air is available on YouTube, on all podcast platforms and Fortinet TV.
