Artificial intelligence is no longer just a tool. It only took a few months for us to reach a point where machines can find weaknesses on their own and attack them on their own. In cyberspace, it is increasingly not people who are facing each other, but algorithms.
AI expert and new rector of the Czech Technical University in Prague Michal Pěchouček is not so worried that artificial intelligence will one day “take control”, as we know it from science fiction. According to him, the real risk is much less noticeable – and all the more dangerous.
Namely, that people will gradually hand over decision-making to algorithms without realizing it. And they will also feel that it is a free choice.
Just a few months ago, the debate about artificial intelligence revolved mainly around its practical use. But today, according to Professor Michal Pěchouček, we are moving further – AI is starting to function as an independent actor in cyberspace.
As an example, he cites the development at Anthropic, which developed the AI Mythos model. “Models are emerging that can automatically analyze the systems of large companies and look for weaknesses in them – without direct human intervention. And that means that the rules of the game are changing completely.”
This is said to fundamentally change the very nature of cybersecurity. “Cybersecurity today is a battle of machine against machine. This is nothing new, I’ve been saying it’s going to happen for fifteen years. And now we’ve really reached that point,” he says.
While people used to be the main line of defense, today their role is changing rapidly. “People are no longer the ones who manually defend everything. They’re more of an innovator – someone who suggests how those defense systems should work.”
What’s crucial is not just where the development is headed, but how quickly it’s happening. “As scientists, we can imagine where it’s going. The only thing we don’t know is the speed. And that’s absolutely key – because if the development is slow, society has time to react. If it’s fast, it won’t.”
That’s why he considers it crucial how technology companies themselves approach the development. “What Anthropic is doing now is interesting – they first gave their AI model to large companies so that they could learn to work with it and prepare a defense when someone wants to use it against them. In fact, they are slowing down the development a little and giving society time. I find that very commendable.”
But one thing is certain, according to Pěchouček, regardless of the speed. Automation in cyberspace will continue to grow and cannot be stopped. “Cyberattacks are expensive today – and the more you automate it, the cheaper it is. And the same goes for defense. So in the end, it’s going to be technology fighting technology.”
Public debate often revolves around the question of whether artificial intelligence will one day “take control” of everything – as we once saw in the movie The Terminator. But Michal Pěchouček also sees other possibilities.
“The fact that artificial intelligence will destroy us one day is a plausible scenario. It could happen. But I think the main danger lies somewhere else entirely,” he says.
According to him, the most likely thing is not that AI itself would threaten humanity, but that people would start using it en masse against each other. “The more immediate danger is the misuse of artificial intelligence by humans against other people.”
In addition, he draws attention to an even less noticeable, all the more fundamental threat – the gradual loss of human autonomy. “The biggest danger is of a completely different nature. It is that humans gradually lose their ability to decide for themselves.”
Artificial intelligence is increasingly entering everyday decisions – from what we buy to who we communicate with.
“You think that today you choose what you will eat, what you will buy, who you will communicate with. But do you really choose? Algorithms are increasingly deciding for you. And you are gradually entering some kind of microworlds that someone is putting together for you. Someone, something. But not you.”
He sees a significant danger in this. “The moment people are so closed in their microworlds, they can be controlled very easily. It used to be science fiction, today we are basically living it.”
According to Pěchouček, it is therefore always necessary to think about technology comprehensively – as a tool, the technology itself, but also about its impact on society and human behavior. “That is why it is necessary for AI to be studied not only by technically oriented scientists, but also by those from other scientific fields. Because AI affects the entire spectrum of human activities.”
There is increasing talk in the public sphere about a “race for general artificial intelligence” between the US and China. According to Pěchouček, this idea is accurate to a certain extent.
Unlike today's systems, general artificial intelligence should not be focused only on narrow tasks, but should be able to solve a wide range of problems in a similar way to humans – from analytical thinking to decision-making in unknown situations. “Yes, it is the milestone. That is what it is about today,” he says bluntly.
At the same time, he points out that this is not a race with one winner. “It is not like a nuclear bomb, where whoever has it first wins. It will not work like that here.” According to him, this also leads to a surprising conclusion: rivals like the US and China will have to join forces at some point. “Ultimately, the powers will have to cooperate to prevent someone from
will misuse it against themselves.”
While the United States and China are investing massively in the development of artificial intelligence, the EU is often perceived – and also ridiculed – as a player that regulates rather than innovates. “That’s a kind of European urban myth that we tell ourselves here, and that’s what we’re doing. It’s not true,” emphasizes the CTU rector.
Europe is said to be lagging behind, but for completely different reasons. “The main problem is that we don’t have a single capital market in Europe,” he explains. Europe simply doesn’t have a strong enough mechanism to finance innovation on a large scale.
“We don’t have the infrastructure that would allow companies to scale in the way that is common in the US,” he says. And the problem is not only systemic, but also cultural. “We don’t have to look only to Brussels. It’s also about how we ourselves behave as investors.”
Using the example of an ordinary European, he shows how the whole mechanism works in practice. “A European – and it doesn’t matter if he’s French, Italian or Czech – makes money. To invest it well, he first buys an apartment. And then invests on NASDAQ, i.e. in American technology companies,” he describes. According to him, the result is paradoxical: European savings largely finance innovation elsewhere. “We pay for innovation ourselves – just not here,” he summarizes.
A systemic decision is needed for change, says Pěchouček. Europe should admit that it cannot succeed in global competition without a strong, respected capital infrastructure.
“We need a real European stock exchange. Let’s not waste time – and money – on the pleasant idea that each country will have its own small stock exchange,” he says openly. According to him, it is precisely this fragmentation that prevents capital from concentrating and effectively supporting the growth of technology companies.
In addition to the economy, according to the rector, the culture of entrepreneurship also plays a role. “Americans are much more willing to take risks. Failure is forgiven there. In Europe, we fear it more,” he adds.
So what is the impact of the often criticized regulation? The professor admits that there is no consensus even among experts, and even in debates with people from the technology sector, he encounters different opinions. But he himself rejects the simplistic view that regulation is automatically an obstacle to innovation.
“We regulate a lot of things – food, transport, security. Thanks to this, we have drinking water flowing from the tap, we have controlled products and a functioning infrastructure,” he reminds us. According to him, regulation is not an ideological tool, but a practical mechanism for protecting society from risks that would otherwise carry too high costs.
According to him, it is therefore strange that part of the debate, especially in the area of artificial intelligence, tries to reject regulation altogether. “Why shouldn’t we regulate artificial intelligence?” he asks. Especially at a time when, according to him, the scientific community has been drawing attention to specific risks for years – from safety to manipulation to impacts on society.
But this does not mean that every regulation is automatically correct. Pěchouček admits that European regulations can sometimes be unnecessarily complex or cumbersome. However, according to him, this is not an argument for their rejection, but rather for their improvement.
“Just because something is not done ideally does not mean that it does not make sense,” he suggests. In short, the European approach is simply different from the American one – and there is a justification for this. “In America, shareholder profits are maximized. In Europe, users are protected more. And because I often feel like a user, I perceive this protection as a relatively significant value.”
According to him, Europe is not standing outside the game, but is playing it in a different way. “The fact that we are not the most competitive region in AI today is a fact. But it does not mean that the European path is wrong.”
The growing influence of technology companies is not only reflected in the economy, but increasingly in politics as well. “There is a gradual growth of technology companies into state power. And in some cases it is starting to get really tight,” he says. He refers to this trend as a “techno-oligarchy.” “Companies developing artificial intelligence are extremely powerful. And they naturally maximize the interests of their shareholders. That is their basic logic.”
At the same time, he points out that it is not fair to lump all companies in one bag. For example, Anthropic, according to him, shows a different approach. “This is a company that is aware of the risks and is trying to work with them. This is not someone who would recklessly push technology out without considering the consequences.”
According to the rector, the difference between the academic and corporate environments is also interesting. “If you look at the top results in AI, there are more of them in America today than in Europe. But if you subtract those that are created in large companies, you suddenly find that European universities are doing very well.”
But then there is a more sensitive question: How much influence can technology companies have on the academic environment? Especially at a time when universities are increasingly collaborating with business and are being funded. Can the money of big technology players indirectly influence the direction of teaching or research?
“An interesting hypothesis. I’m not saying it’s not possible – but I haven’t actually seen any blatant cases of technology companies systematically influencing universities,” says Pěchouček.
He cites a rather hypothetical situation as an illustration: “I can imagine
that a big sponsor would come and say: ‘We don’t want this topic to be researched here.’ For example, research on the risks of artificial intelligence. But I don’t think it would be worth it for them – it would discredit them. And universities must have codes of ethics that clearly define cooperation. And most importantly, they must remain independent – from companies and politicians.”
According to the rector, the broader role of the university in society is also essential. “The university should be a safe place for open debate. Even uncomfortable topics should be addressed there – regardless of who is the sponsor or who is in power.”
One of Michal Pěchouček’s main goals is also related to this: he wants to change the functioning of the Czech Technical University in Prague so that it is much more closely connected to society. “The biggest risk is that the university will close in on itself and cease to be relevant.” He rejects the idea of the school as an isolated “ivory tower.”
“Just because we’ve been here for three hundred years doesn’t mean we’ll be here for another three hundred years. Things are changing so fast these days that ‘leaving something out’ can mean you become irrelevant and out of the game.”
According to him, the university must actively respond to changes in society and technology. “The school must understand what’s happening outside – in industry, in society – and it must reflect that in teaching, research and cooperation.”
He spoke similarly in his inaugural speech when he took up the position of rector. The ceremony was attended by personalities from across segments – including top business. This is rather unusual in the Czech environment – similar events are traditionally attended mainly by the academic community.
According to him, however, the composition of the guests illustrates well where he wants to move the university: towards much closer cooperation with the private sector.
According to the reactions from practice, he has made very good progress. “CTU has a new rector who is not afraid to dream, he has been holding on to his dreams for thirty years and is gradually fulfilling them. I believe that the next dream – to move the university at least among the top European universities – will become a reality,” said Hana Součková, CEO of the Czech branch of SAP, on site.
Alexandra Kala, CEO of Profimed, also speaks similarly: “He has exactly the profile that can move CTU towards greater dynamism and modernity. His election is extremely positive news for Czech science and education – and therefore also for the economy.” However, the new rector rather lowers his eyes when reading these words. “It sounds great. But it is also quite binding. It is a huge responsibility,” he says.
How does he want to fulfill his plans? One of the key ways is to provide more significant support for startups. These days, these are often created outside universities, as purely entrepreneurial projects, but according to him, the academic environment should be a natural part of them. “We want it to be normal for a professor or a doctoral student to start a startup. Today, it’s still not a given,” he says.
At the same time, he openly admits that the current system is not very conducive to this. “For the head of a department, a startup is often a problem. He has someone who teaches well, he needs them for teaching – and suddenly that person leaves to start a company. And the head has a hole. The natural reaction is then: it’s better not to create that startup.”
It is precisely this way of thinking that the rector wants to change. “We have to turn the logic around. Society doesn’t just expect us to teach. It expects three things: teaching, research and technology transfer. And a startup is exactly that transfer,” he explains.
According to him, universities should actively support the creation of new companies, not hinder them due to short-term operational problems. “When I see that someone has the potential to build a company, I should want it to be created – and ideally, for more to be created. Not hinder them because I miss them in teaching.”
But the Czech environment has a broader problem: a lack of courage to take risks. “Researchers often bet on safety. They submit projects that have a better chance of passing, but are not as ambitious,” he says.
According to him, the situation in industry is even more critical. “Czech industry is even more stuck in the wall than academia. Companies want quick solutions – something that will be ready in three months.” According to him, this creates a fundamental mismatch.
“The university is set up to solve complex problems with a horizon of five to ten years. The company wants to fix a specific thing right away. The two worlds often do not understand each other at all.” Nevertheless, he sees a great opportunity in this. “We have strong companies here, capable people. We could easily be Switzerland in innovation. But we must want to solve big things, not just small adjustments.”
According to him, however, universities should not primarily support startups for the sake of money. “We don’t do startups to get rich. We do it to meet society’s expectations. They expect us to do more than just teach and research, but also to have a real impact.”
That’s why he advocates a so-called founder-friendly approach. “When a university takes a large stake in a startup, it often kills it. It reduces its chances of success.” His logic is simple: “It’s better to have a small percentage in a successful company than a large percentage in an unsuccessful one.” He looks for inspiration in top European institutions such as ETH Zurich or the Technical University of Munich. “Those schools don’t want to own the startup. They want it to succeed. And if it succeeds, they still get something out of it – but that’s not the primary goal.”
Michal Pěchouček is called a visionary, especially thanks to his expertise in the field of AI. He himself is a bit resistant to this label, describing himself more soberly: "I am a person who thinks a lot about the future. I look ahead, I put things together." But he admits that even that has its limits. "Sometimes I get it right, sometimes not. But more often I do. And because of that, they are a bit allergic to me at home," he adds with a smile. "But I guess that's part of it."