Timotheus Höttges: »KI hilft uns mit unserer Überforderung«

Haben Sie kurz Zeit?

In this episode of 'Haben Sie kurz Zeit?', Rainer Esser interviews Deutsche Telekom CEO Timotheus Höttges about the intersection of telecomm

Key takeaways

• Timotheus Höttges views AI as a critical invention to help humanity master the 'overwhelm' of modern complexity.
• Deutsche Telekom manages between 60,000 and 70,000 network attacks per minute using automated machine intelligence.

Main topics

• Artificial Intelligence as a tool for complexity
• Digital sovereignty and European infrastructure

Notable quotes

AI helps us to become the master of our overwhelm.

Conclusion

The episode offers a profound look at how legacy industrial giants are

Transcript preview

Speaker 1 (0:10) This episode was recorded on 5.8.2026. Speaker 5 (0:23) Dear listeners, welcome to the Zeit Podcast. Have a short break. My name is Rainer Esser, I am Executive Advisor at Holzbrink, the time's supervisory board, and 27 years ago, CEO of Zeit. In my podcast, I talk to bosses of big companies. It's not about strategies and the numbers, but about the people behind the role, what drives him, what moves him, where he wants to go, what he finds exciting and wants to pass on. My guest today is Tim Höttges, the CEO of Deutsche Telekom. Speaker 4 (1:09) Thank you Speaker 3 (1:10) very Speaker 4 (1:10) much. Tim Höttges, that you are here. Good afternoon, Herr Esser and all listeners. Speaker 5 (1:15) For those of you who don't Speaker 4 (1:18) know Speaker 5 (1:18) Tim Höttges too well, a short break. He is one of the few people who you have to say the professional consultant could have saved the appointment. Tim Höttges knew pretty well as an abiturient what was to become of him. Others wrote in the Abitur newspaper, probably doctor, journalist or something with media, Tim Höttge wrote, chairman with a salary of 100,000 Mark. Speaker 5 (1:54) Das mit dem Vorstandsvorsitzenden hat geklappt. Bei den 100.000 Mark ist die Sache ein bisschen aus dem Ruder gelaufen. Seine kaufmännische Ausbildung begann er nicht im Vorstand, sondern schon auf dem Wochenmarkt. Mit 16 verkaufte er dort morgens vor der Schule und samstags Butter, Milch und Käse. Heute verkauft er Glasfaser, Mobilfunk und Magenta. Im Grunde dasselbe Geschäft, nur die Kühlkette ist etwas einfacher. Since 2014 he has been leading Deutsche Telekom. Now he is one of the oldest heads of staff at DAX and is the biggest with 1,93. He lives privately with his wife in Bad Godesberg, has two sons, and calls his marriage a life-time subscription. Together with his wife he is engaged in the Bürgerstiftung in Bad Godesberg. He promotes art, is interested in artificial intelligence and is committed to the common good. A telecom boss who not only networks people, but also wants them to talk a lot and properly afterwards. His contract has just been extended again. The man who wants to become a vice president as an abiturient, seems to be not finished at the best age. Speaker 5 (3:26) Deshalb möchte ich heute vor allem eines wissen. Was träumt man eigentlich noch, wenn man die Träume aus der eigenen Abizeitung längst erfüllt hat? Ich freue mich auf das Gespräch mit Ihnen, Herr Höttges. Ich auch. Bevor wir beginnen, können Sie uns kurz sagen, wo wir hier eigentlich sind, den Hörerinnen und Hörern. Es ist im Augenblick ein bisschen zugezogen, aber wenn die Vorhänge auf wären. Was würden wir sehen, wenn wir sehen könnten? Werbung. Speaker 3 (4:24) Find out more about classics, casuals and sneakers for ladies and gentlemen at the official webshop at florisfranbommel.de. Speaker 4 (4:38) Yes, welcome to the living room of Deutsche Telekom. This is the central one here. We are on the fifth floor. We are in a converted studio. It's all improvised. It's a normal office that is now used as a podcast studio. We deliberately avoided large investments, but organized this podcast with simple means, because ultimately Telekom is a very efficient company. The building was once planned as a public building and has gone through many, many times, as we as a company have too. And that's why it's more like a living room today, which is designed Speaker 5 (5:17) very Speaker 4 (5:17) loosely, easily. Maybe a little more inspiring, more like a valet office space and less like a office room. Speaker 5 (5:29) When you arrive here at Deutsche Telekom, here at the main office, There is a dancing elephant in the office. It is dancing on its back. What does this elephant mean to you? Is there any connection? Speaker 4 (5:45) For me, there is a very big emotional connection. I try to create it with the company. And we succeed very well. The elephant is dancing on its back. already by definition impossible. Speaker 1 (6:01) An Speaker 4 (6:02) object by the world-famous artist Miguel Barceló, Spanish sculptor, who makes the impossible possible. At our 30th anniversary celebration, the Telekom thought we want to leave something behind, not just a party for the city or our customers, not just celebrate ourselves, how great we did all this, but maybe also create a monument in the best sense of the word, where we say What was the generation that is still at work for? And what is Telekom for Speaker 3 (6:35) in Speaker 4 (6:35) the future? And every day people come in and out of our company. Speaker 3 (6:39) And Speaker 4 (6:40) when everyone gets an impulse and says, I'm trying to make the impossible possible, to bring the elephant to dance, then that has an enormous power to be courageous, to reinvent oneself again and again, to continue to transform. Because stillness is dead. And that applies to companies in digitalization and telecommunications even more. And that's why the elephant for us is a reminder of what we have achieved in recent years, but also a monument for those who will come and go in Speaker 5 (7:10) the future. You often go to the USA. You were in Silicon Valley recently. You were in Seattle, in Bellevue, with your American daughter, T-Mobile US. And when you came back, you said that AI is the greatest gift of our time. What does this gift contain? Why is it such a great gift for you and for Telekom and for all of us? Speaker 4 (7:39) So first of all, I think any form of too much protectionism, too much to be turned inside out and to travel too little and to make themselves smart for one of the greatest dangers of our time. Who says that we are on the right side of history? Let's remember the conflict between East and West. Both sides were convinced that they were on the right side of history. And afterwards you can discuss who was on the right side. And I'm experiencing exactly the same thing again now. I consciously take the time to travel to Asia once a year, usually also to travel to Israel at least once, in the face of the hindrance of the innovations that are possible there. And then of course to visit the American ecosystem, especially the Valley, where the AI has also been invented, where AI has actually been commercialized through OpenAI, but actually much earlier through the GPT development of Google and the fundamentals that have been laid down in the academy by Stanford. Speaker 4 (8:43) AI helps us to become the master of our overwhelm. Whenever a person has reached an overwhelm, he always invents something. At some point it was the physical subject, ordering the fields, then the wheel had Speaker 5 (8:58) to be Speaker 4 (8:58) invented. At some point it was the knowledge transmission, then the book was invented. And at some point the computer was invented, because the complexity has simply become so great for the work-related processes. or the steam engine that enables industrialized production. We live in a time that has enormous demands on scientific development. We have a time that has enormous demands on taking away our routines. At the same time, we have a complexity of over 7 billion people who have to be organized on this planet. And in this complexity we invent a tool, again a tool, was uns helfen wird, mit dieser Komplexität besser umzugehen. Es wird nicht alle lösen, Themen, die alle Probleme lösen, aber es wird helfen, in dieser enormen Herausforderung der Menschheit ein Stück weit Herr zu werden. Und sie wird unser Leben erleichtern. Sie wird uns neue Spielräume geben für Kreativität. Sie wird uns lästige Arbeiten abnehmen. Sie wird uns hoffentlich, wenn wir nicht faul werden, und viele werden faul werden, wird sie uns auch schlauer machen. And in the background, it will also help us to pick up deficits. Take, for example, the Germans who are not good at software. We have lost this software and digitization. Not from the academy, they can do that. But from the industrialization of software. That started 15, 20 years ago. And now we have the chance to pick up many of these deficits with artificial intelligence. So again, a gift. And I could talk an hour longer now. I am an absolute technology optimist in the relationship, knowing that everything is a poison. But the chances for me clearly exceed the risks. Speaker 5 (10:45) How can you use AI here in the company to solve the complexity at Telekom and to facilitate the work of your employees and especially your customers? Speaker 4 (10:59) At the Telekom, I stopped counting, there are currently over 500 projects running at the moment. And the main Speaker 5 (11:04) projects? And Speaker 4 (11:05) the projects are in all areas. That's also what's fascinating about AI. Call me one area of life where AI doesn't have an influence in any way. The preparation of your podcast about AI, the distribution in the future on the net about AI. Everything is influenced by AI and that applies to the business of the Telekom as well. The first topic is the networks. The networks are the basis for what Telekom produces. And our networks are attacked from the outside. We currently have 60,000 to 70,000 attacks per minute. Per minute on our infrastructure. Speaker 5 (11:44) Where do they come from? They Speaker 4 (11:45) come through software that is on the network. This is regular software. We already know most of it. but also always new types of attacks and vectors that attack our network from different areas Speaker 5 (11:57) of Speaker 4 (11:57) the world. And they want to take Speaker 5 (11:59) data. They want to take Speaker 4 (12:00) data, they want to identify weaknesses. If we try to control all this through humans, it wouldn't work. We need machines for this, and the more intelligent these machines are, the better. Secondly, a practical example. You are on the road, drive into a traffic jam. This traffic jam is not predictable. How does the network react to it? Today, you have a certain spectrum of radio that is assigned to every customer. Suddenly, there are many more people in this cell than before. The network is dynamic, it now divides the radio cell. You get a little less bandwidth, but everyone stays in the network. We can't catch all these anomalies in our infrastructure so quickly today. That's what artificial intelligence does today. It sees that a lot of traffic is currently emerging. I have to do something and controls the antenna individually. auf dem Mast, auf dem Hausdach, im Kirchturm, entsprechend um. Dann haben wir den Bereich Sales and Service, der gesamte Vertrieb und die Kundenbetreuung. Früher hatten wir das Thema lange Wartezeiten. Speaker 5 (13:02) Heute gibt es FragMagenta. Speaker 4 (13:04) Heute gibt es FragMagenta. Unser Chatbot, den Sie benutzen, wenn Sie zum Beispiel mit uns im Chat sind, beantwortet heute über 60 % aller Anfragen im Erstkontakt, ohne dass der Kunde ein zweites Mal anfragen muss. And that's fully automated. Now I say 60 % is a good start. Of course I want over 80%. But even our people who work in the company can't answer all questions in the first call. Now we're going to do the whole thing in language. That means they will also use language. Let's call it Roboter. But that's a topic that people don't like. These are assistants, language assistants, who are trained with AI. Speaker 4 (13:46) The first wave, so to speak, work. And if that's not enough, there's always a person behind it. There will always be the human aspect at Telekom, but of course our system also relieves gravity. Then there is the entire administrative area. Take routine activities in the bookkeeping or a bill that is available digitally today. It will be completely available to everyone. über die Compliance-Abteilung, über die Revision, über die Buchhaltung bis in entsprechend zur Rücküberweisung an den Mitarbeitern völlig über Agents automatisiert durchgeführt. Das ist heute noch nicht live, aber das wird in der Zukunft einer der normalen Prozesse sein, wie wir unsere Workflows, wie wir das hier intern nennen, unsere Arbeitsprozesse reorganisieren werden. Und vielleicht das letzte Beispiel, auch was Praktisches. For the listeners, we have Ask T, also Frage das T. Every employee has on his computer the entire knowledge of Deutsche Telekom in access. That means we have bundled our entire data in a cloud infrastructure and the employees can ask everything via a AI assistant. If we now, for example, say how many antennas do we have in Greece, then the system will answer exactly the same as how many holidays I still have today. That means we bundle the company's knowledge over a database and the employees can access the knowledge in this way. In the past, they used websites, had to register again and again, then they didn't get the questions or had to speak to personal assistants who were not available. All of this will soon be supported by, or is already supported today, by a corresponding AI assistant. Speaker 5 (15:31) So for the company a great relief and you They have a lot of projects here to improve the processes in the house for the customer and for the employees. What about new services that you offer to companies? Telekom has historically been Speaker 4 (15:55) made Speaker 5 (15:55) of masks and fiberglass or copper cables. And now they also go to computer centers. Now they are going to a new field. where there are already a few other very large ones. Speaker 4 (16:08) So first of all, you see the diversity where AI is used today and AI is needed, the challenge. And if you ask me today, what is an important part of your day job, then it is AI. I would say that I was dealing with 30 % of questions every day about how we get artificial intelligence to our customers, how we can better serve customers. with AI and how we can better organize our work processes with AI. The question is, what do we do for our customers? First of all, we can develop software and products much faster. If you open the Magenta app today, for example, my Magenta and take a look at the app, you will find a lot of offers that are individually responsible for our customers. You don't have to earn loyalty with us, but every customer gets the so-called advantages of Magenta offers through our relationship with their mobile phone tariff. These are tickets, these are the corresponding products that we can offer, these are travel support or other topics that are Speaker 5 (17:12) individualized by Speaker 4 (17:12) AI. Speaker 4 (17:15) to Speaker 5 (17:15) be controlled on Speaker 4 (17:15) the respective app. The AI decides which offer you get. And you can see that in the app at the back. We are currently implementing financial services. In the past it was very complicated. Today we can develop the SEFT ourselves. We have introduced a Paysy credit card and payment function. And in America we are currently one of the largest credit card and financial service providers this year. with the T-Mobile service. A huge business, three-digit million, which we earn with it. So these are also new services. And then, of course, comes the entire area of business customers. The Telekom has today, maybe also for the thing, we make about 110, 115 billion in sales. About 15 billion of this we make today in Europe only with business customers. And of course we are very important for business customers. And these business customers trust us, We also offer cloud services, security solutions and business applications, especially through DT Systems. And today, nothing is possible without AI. We are very intensive. And now you speak as the last one, the new business field. What is new? The Speaker 5 (18:30) computer center Speaker 4 (18:31) with Speaker 5 (18:31) NVIDIA. Speaker 4 (18:33) We have also said that we understand infrastructure. The German economy has 300 billion of bound capital. That means we know how to deal with long-term capital-intensive goods, with long-term business case, just like with fiberglass or with mobile phone antennas that are built. They don't amortize in a year. Let's build a modern data center. Today we operate 50 % of these high-performance GPUs. These are the graphic processors that are used today for artificial intelligence. We operate them in Germany today and we have now also made a billion