Volker Wissing: ‘AI will determine everyday life, competitiveness and safety’

How successful is the German government's digital policy? c't met Volker Wissing, Federal Minister for Digital Affairs and Transport, in Berlin to draw an interim conclusion after three years of the traffic light coalition. In the interview, Wissing emphasizes the progress made in expanding the network, while keeping all options open regarding the possible exclusion of Chinese suppliers. For Wissing, there is no doubt about the importance that AI will have in our lives. He wants to have oversight and set up a certification system – and has already chosen his favorite for an AI supervisory authority.

Mr Wissing, the traffic light promise of progress was also made for digitalization. How much of this has been delivered?

We have made great progress with the digital infrastructure. 97% coverage with 4G, a good 91% with 5G, 50% more fiber optic connections within a year. The gigabit strategy is flying. We have made progress in implementing the digital strategy: in healthcare policy, the e-prescription, the electronic patient file, the Health Data Utilization Act and thus securing the future of Germany as a location for pharmaceutical and medical research. We have i-KFZ and the introduction of digital registration, the e-ID, which is the key to communicating digitally with citizens.

You can see that things are happening everywhere. We have also pushed ahead with digitalization in the railroads, the modernization of public transport, but have not yet implemented it in all structures.

The e-ID is a so-called lever project of the digital strategy: a prerequisite for successful digitalization. Little has happened. Does this mean that the digital strategy can no longer work?

No. The Ministry of the Interior is responsible for ensuring that a user-friendly digital identity is created - and the BMI is working on this, also in terms of integrating it into a pan-European concept. The e-ID is a very important lever: if the state cannot establish identity digitally, there will only ever be a partially digital administration.

When funds became scarce, this was the first area where savings were made. Digitization was supposed to be a priority at the beginning of the legislative period, the idea was a digital budget as a guarantee that central projects would be guaranteed - could that still come after all?

So far, we have shown that it is possible without a digital budget because individual departments have reallocated and made sufficient budget funds available for their digital projects.

There is a greater need for funding for the e-ID, but this can also be solved by prioritizing again. It will only be possible to say whether a digital budget is necessary to implement the e-ID once the budget discussions are over.

The AI regulation has been passed and will come into force. It has been criticized from many directions. Is Germany ready for the age of AI?

Artificial intelligence will determine our everyday lives, our competitiveness and our security to such an extent that we need to have this technology in our own hands. We must not make any mistakes here and lose sovereignty as a result.

That is why we must not hinder the development of AI in terms of regulation, but must support it with innovation-friendly regulations. This requires courage. Because today we don't know everything that AI can do or will be able to do in the near future. The developers themselves are humble in the face of the possibilities. But if the state intervenes to slow things down, creative minds will look for other locations. We need to experiment, developers need freedom.

Do you see companies in Germany being willing to do this?

Yes! I am delighted to see that SMEs are investing in Germany. And I think that's very good, because it also safeguards our values. For German entrepreneurs, this is not just pure business, they are also pursuing ethical goals.

Nevertheless, as a state we have to make sure that we can understand what is happening. Take the example of the banking crisis. It was said that the state did not need to intervene, that the banks would regulate themselves in competition. As a result, the financial markets created a system in which no one was forced to take responsibility.

Is that what you fear with AI?

No, but we have to learn from that time. After the financial crisis, people asked: Why didn't the state notice? Pretty much everything had triple-A ratings, people thought they had a false sense of security and believed there was no risk. The system had slipped away from the state. The speed at which the financial markets were unleashed and the speed at which supervisory and control structures were established were completely asymmetrical. Then it was said: How is the lawyer at the Federal Financial Supervisory Authority supposed to control Goldman Sachs? But that was precisely the task - and it was suggested to the public that this was taking place.

How can we ensure that this does not happen again with AI?

By not just watching, but participating ourselves. And say: we want to understand, we want to have oversight, we want to set up a certification system. But there is one big difference to the banking crisis: this time, the state is not just a supervisor, but a user of the technology itself.

The German AI supervisory authority is currently being sought. Who will it be?

The Federal Network Agency.

Old disputes are also continuing in the traffic lights: chat control, data retention and the right to encryption. The latter is one of your proposals. Does it have a chance in this legislative period?

Yes, a society based on the division of labor can only be successful with a good culture of trust. This means that communication can be confidential. Encrypted communication is becoming increasingly important for the entire economy. There are more and more risks due to geopolitical changes, while at the same time dependency on data and digital communication is increasing. But it's not just an economic issue. If we restrict our private communication, we lose what is the greatest value of our society: freedom.

The debate about Chinese telecommunications suppliers is also related to this. What is your current view?

I have said from the outset that if we are going to discuss such an issue, we need to have the facts. What exactly are the security risks? I mean...

Do you have the facts now?

We have a clear insight into the technology installed in the network. And we can assess whether and to what extent there is a need for action and when. At the end of the day, politics is about weighing up concrete risks, risk scenarios and other interests that need to be considered. The Federal Government will continue to deliberate and come to a common opinion. However, this is not a topic for public discussion. The weighing up of national security interests cannot be made transparent – but the final result can.

But it's not just about 5G networks. Components from countries of dubious reliability can be found in many areas. Is this being discussed for all critical infrastructures every time?

We have to keep an eye on security issues at all times. A 5G network with real-time applications, for example in autonomous driving, has entirely different security requirements than the previous network.

That's why we always have to look to the future: are today's networks equipped for future risks? If not, you have to react and adapt things.

And that applies from antennas to cloud services to...

But yes! Companies are doing the same. If business models are increasingly data-based, this means we need to focus more on cybersecurity issues. Can our business model be hacked? What does data leakage mean for us? What are the consequences if our system is down for an hour? Companies are asking themselves these questions and creating their own clouds and cybersecurity systems. The same applies to us: We have to deal with it.

Once we have implemented the digital European Train Control System (ETCS) on the railroads, the security requirements will be thoroughly different from today, when we are basically still using the signal control system from the last century. Of course, you have to ask the question: Will our network then be sufficient? With all its components? This is also about the supply chain. And not just the question: Is a supplier trustworthy? It's also about whether subsequent deliveries are guaranteed.

We have seen what it means when supply chains are disrupted. These are also security issues. We can't say: unfortunately we can't offer train services for the next few weeks, we can't get the spare parts. That must not happen.

So sovereignty is the solution?

When it comes to what we need, such as semiconductors, we are always focused on sovereignty and want independence. But the truth is that in a world based on the division of labor, the complete sovereignty of a society cannot be guaranteed. Because we cannot produce everything we need ourselves.

That's why we have to manage risk and not make ourselves dependent on individuals. That is our task, and it is big enough.

(vbr)