News

An interview with SAYTEC CEO Selim Kuzu by Markus Plettendorff – 18 December 2025

The international court in The Hague plans to replace Microsoft products with a German solution. The order for American F-35 fighter jets is viewed critically because there is hardly any access to their software. Do you understand these concerns? How important is digital sovereignty?

Selim Kuzu: Digital sovereignty is of central importance for Germany and Europe – strategically, economically, politically and socially.

What does digital sovereignty actually mean?

Selim Kuzu: Digital sovereignty is not a luxury project but a prerequisite for Germany’s competitiveness, security and democracy. It will determine whether Europe shapes the digital future or merely uses it.

Digital sovereignty describes the ability of states, businesses and citizens to:

• use technologies on their own terms,
• control their own IT infrastructures and data,
• and avoid dependence on foreign providers or platforms.

In short, it is about independence, control and the power to shape the digital world.

Why is it so important for Germany and Europe?

1. Economic independence

• Large parts of the digital infrastructure (cloud, chips, software) come from the USA or Asia.
• This dependence creates risks for supply chains, pricing and technological standards.
• European alternatives such as SAYTEC, GAIA-X, the European Cloud Federation or the Chips Act are intended to counter this.

Goal: building our own innovation and value chains.

2. Security and data protection

• European citizens’ data is often stored on servers outside the EU, which creates legal and security-policy risks.
• Digital sovereignty protects against cyberattacks, espionage and dependence on foreign security standards. Access to this data could also be blocked – in theory, or even technically.
• It strengthens the trustworthiness of critical infrastructure (e.g. energy, healthcare, data centres).

3. Political and strategic autonomy

• In geopolitical crises (e.g. trade disputes, sanctions), technological dependence can become an instrument of power.
• Europe wants to avoid becoming a “digital protectorate” of other powers.
• Digital sovereignty strengthens its negotiating position in international relations.

4. Social and cultural self-determination

• Digital platforms influence opinion-forming, education and democracy.
• European solutions could promote transparency, data protection, ethical standards and new value systems based on European values.

It is also about digital identity and democracy in the 21st century.

How can this sovereignty be ensured? Do companies have to build new, complex IT structures of their own? Or are there (competitive) alternatives to the big cloud solutions from AWS, Google and the like?

First of all, there has to be awareness – in other words, a new mindset. Germany is still far too comfortable: it sees neither the risks nor the opportunities.

Challenges

• High investment requirements in research, semiconductors, cloud and AI.
• A lack of economies of scale compared with US or Chinese tech giants.
• Fragmentation of regulations and strategies across EU member states.

Strategic opportunities

• Building a European data economy (“data spaces” in industry, healthcare and energy).
• Promoting open-source technologies and European cloud solutions.
• Strengthening SMEs and Industry 4.0 through trustworthy digital infrastructure.

Let’s talk about AI – generative AI and AI agents, which are now used both to launch attacks and to defend against them. To what extent is AI already integrated into companies’ security architecture?

Selim Kuzu: A very good and highly topical question – and one that goes to the heart of digital transformation.

Artificial intelligence (AI) is already a central component of modern security architectures in companies, although to very different depths depending on the industry, company size and security maturity. In Germany, unfortunately, it is still hardly used.

Here is an overview of how far AI is currently integrated into IT security architectures and where things are heading:

1. Status quo: where AI is already used today

a) Threat detection and anomaly analysis

• AI models (often based on machine learning) monitor log files, network traffic and user behaviour.
• They detect unusual activities, patterns or deviations that indicate cyberattacks, malware or insider threats.

Examples:

• SIEM systems (security information and event management) such as Splunk, Microsoft Sentinel or IBM QRadar use machine learning for pattern recognition.
• EDR/XDR systems (endpoint/extended detection and response) – e.g. from CrowdStrike, Palo Alto Networks or Sophos – rely heavily on AI to detect attacks.

Already standard today in large companies and critical infrastructure.

b) Automated response and incident response

• AI systems help security operations centres (SOCs) to prioritise alerts and initiate countermeasures automatically.
• In the event of an attack, an infected host can, for example, be isolated automatically.
• AI reduces false alarms (false positives) and increases analysts’ efficiency.

Advanced companies use AI-based orchestration (SOAR).

c) Identity and access management (IAM)

• AI detects anomalous login attempts, unusual access times or new devices in order to uncover possible compromises.
• Adaptive authentication systems (e.g. Microsoft Entra, Okta) use AI to control risk-based multi-factor authentication.
• Already in use for cloud services and at larger companies.

d) Phishing, fraud and e-mail protection

• AI-based filters analyse text, metadata and sender behaviour to detect phishing, social engineering and deepfake attacks.
• Major providers such as Google, Proofpoint or Barracuda use generative AI for real-time assessment.

Already commonplace in e-mail and collaboration platforms.

e) Vulnerability management and patch intelligence (general)

• AI helps to prioritise vulnerabilities based on risk analyses, the likelihood of exploitation and the company context.
• Tools such as Tenable, Qualys or Rapid7 use AI to assess the “real-world exposure”.

Increasingly used in large corporations and regulated industries.

2. Limitations and challenges

• The black-box problem: many AI systems are difficult to understand, which makes compliance and audits harder.
• Bias and wrong decisions: AI can learn faulty patterns, e.g. from unusual user behaviour in legitimate scenarios.
• Lack of expertise: security teams need to learn how to understand, train and monitor AI models.
• Data protection and GDPR: in Europe in particular, the handling of security-relevant data is highly regulated.

3. Future trend: from AI-assisted to AI-driven

a) Autonomous security systems

• Future architectures will integrate generative AI that dynamically adapts security policies and mitigates vulnerabilities on its own.
• Examples: “self-healing networks” or “autonomous SOCs”.

b) AI as protection against AI attacks

• With the rise of generative AI attacks (deepfake phishing, automated exploit generation), defensive AI is becoming even more decisive.

c) Integration into Zero Trust architectures

• AI takes on central roles in the continuous risk analysis of identities, devices and applications.

But there is still the human factor – people can easily fall for perfectly crafted phishing e-mails and the like. To what extent is this (sufficiently) taken into account in IT security?

Selim Kuzu: A very good follow-up question – and one that touches the sore point of every security architecture. Because, despite all the technology, one thing still holds true:

People are the weakest, but also the most decisive, link in the IT security chain.

Here is a well-founded assessment of how well IT security is prepared for human error today, using phishing and social engineering as examples:

1. The reality: 80–90% of all security incidents start with human error

• According to IBM, Verizon and the BSI (German Federal Office for Information Security), most cyberattacks can be traced back to human error – usually via phishing e-mails, weak passwords or careless clicks.
• No technical system can rule this out completely.
• As a result, many companies’ security architecture has evolved from purely technical defence towards “human-centred security”.

2. How security architectures are prepared for human error today

a) AI-supported detection of phishing and social engineering

• AI analyses the language, tone, style and sender behaviour of e-mails.
• Suspicious messages are automatically flagged or blocked.
• Systems such as Microsoft Defender 365, Proofpoint or Barracuda Sentinel combine machine learning with global threat intelligence data.

Technically very advanced – but not 100% protection.

b) Multi-factor authentication and Zero Trust

• Even if an employee falls for a phishing e-mail, multi-factor authentication (MFA) prevents attackers from gaining direct access.
• “Zero Trust” models assume that any access may be compromised – and require continuous verification.

Significantly reduces the consequences of human error.

c) Security awareness and behavioural training

• Companies increasingly rely on continuous training, phishing simulations and awareness campaigns.
• The goal: fostering reflexive caution, not just imparting knowledge.
• Modern training is gamified, data-driven and AI-supported (e.g. KnowBe4, SoSafe, Lucy).

Effectiveness: high, but dependent on corporate culture.

d) Automated damage limitation

• If a click does happen, automated response systems detect unusual activity (e.g. sudden data exfiltration, unusual logins).
• Systems isolate affected accounts or devices in real time.

Minimises the impact of human error.

e) “Human firewall” – integrating people into the architecture

• Modern architectures see employees as part of the security ecosystem, not as a source of risk.
• Users can report suspicious e-mails with a single click; these are then analysed automatically by AI.
• User feedback in turn improves the learning models.

People and machines are increasingly working in symbiosis.

3. Limitations and weak points

Despite all the technology, the following remains true:

• Stress, time pressure and routine lead to rash actions.
• Deepfake attacks (e.g. deceptively realistic voices or videos) make social engineering even more sophisticated.
• A lack of security culture in many organisations means that employees do not dare to report mistakes immediately.

Technology can catch mistakes, but it cannot make up for being caught off guard.

Thank you very much for your concise yet clear description of the situation and the solutions!

November 2025

The geopolitical, economic and technological developments of recent years have shown impressively how essential digital sovereignty, resilience and security are for Germany as a business location. At the heart of these challenges are high-performance, highly secure IT infrastructures and reliable cybersecurity concepts and solutions. After all, digital transformation brings considerable dependencies with it. Particularly with regard to cloud services, data processing and critical infrastructure, outside influence – above all from the USA – is substantial. To ensure digital sovereignty, companies and institutions must keep control of their data within their own country – not only for data protection reasons, but also to preserve their strategic independence.

Guest article by SAYTEC CEO Selim Kuzu in Handelsblatt

By Axel Höpner

Better than US competitors such as Hewlett-Packard or IBM? German IT specialist SAYTEC sees itself as the technology leader – and counts intelligence services among its customers.

Taking part in this year’s CloudFest gave us the opportunity to follow the latest trends and innovations in cloud technology at first hand.

With more than 400 participating companies and over 20,000 visitors, this global event is one of the most important gatherings for cloud and digital technologies in Europe and worldwide.

In an interview with our global R&D partner Solidigm, we at SAYTEC explained how our strategic collaboration enables us to secure IT infrastructures with the latest cloud, secure remote access and storage solutions.

We formed many new strategic partnerships and had the chance to examine new products and technologies on site. Direct contact with key decision-makers in the industry also gave us inspiration for developing specific solutions for the European market.

Thanks to our R&D partnership with Solidigm, which began at CloudFest, we are continuing our successful collaboration with TechArena. Following on from our recently released podcast, our joint blog post “Insights into the IT Infrastructure Revolution with SAYTEC” has now been published on TechArena’s social media channels.

How industrial companies protect their networks: IoT and machine-to-machine communication – highly secure.

sayTRUST ZeroTRUST put to the test: how to work highly securely from anywhere.

Digitalisation, authentication and remote work have long been a reality for every company. But at a high price: cyberattacks cause €148 billion in damage to the German economy every year. According to a Bitkom study, 84% of all German companies have already been affected – a sign of how urgent it is to address IT security.

Workation, kitchen, Wi-Fi, hotspot, office – office staff, technicians, digital nomads and coworking-space users log in to company networks from different places using different devices. But is the person logging in really who they claim to be? And is the network they are using secure?

A holistic view of IT infrastructure

Work highly securely from anywhere with sayTRUST VPSC

With the all-in-one platform sayFUSE, founder Yakup Saygin has created a solution that offers both simplicity and security.

Digitalisation, authentication and remote work have long been a reality for every company. But at a high price: cyberattacks cause €148 billion in damage to the German economy every year.

Office staff, technicians, digital nomads and coworking-space users log in to company networks from different places using different devices. But is the person logging in really who they claim to be? And is the network they are using secure?

sayTRUST only allows communication after an eight-stage security procedure has been completed successfully – and it does so from encrypted memory. This digital gatekeeper intercepts unauthorised programs or malware in advance and strictly checks the individual permissions of anyone requesting entry.

Simplifying and securing remote work: the installation-free Zero Trust security solution sayTRUST VPSC from developer SAYTEC enables secure access to confidential networks from any location. An eight-stage security procedure checks the user’s biometrics, PIN entry, fingerprint and personal key identity and grants access to individual applications – making self-managed monitoring and updating of end devices a thing of the past.

Digitalisation, authentication and remote work have long been a reality for every company. But at a high price: cyberattacks cause €148 billion in damage to the German economy every year.

Pages 50–51: The ideal, secure network infrastructure

IT chaos slows down processes and puts obstacles in the way of cybersecurity. With the all-in-one platform sayFUSE, founder Yakup Saygin has created an offering designed to provide simplicity and security.

Tangled cables, hardware chaos, a labyrinth of software – anyone who wants to equip their company with IT infrastructure and security solutions can quickly lose track. IT chaos slows down processes and puts obstacles in the way of cybersecurity. According to IBM’s Cyber Security Intelligence Index, 90% of all cyberattacks are caused by human error. Munich-based IT developer SAYTEC takes a holistic view of IT infrastructures and technologies.

Anyone who hosts data needs a virtual infrastructure that protects against outages and data theft and can process as much data as possible. When companies piece together IT solutions from numerous different manufacturers, the risk grows that software and hardware will not be compatible with one another.

The ongoing state of emergency in which German healthcare facilities have been operating due to the pandemic has accelerated the search for new, mostly digital working models. To reduce the risk of infection in the particularly vulnerable hospital environment, working from home has become a genuine solution wherever direct contact is not essential.

Transferring large volumes of data is now an everyday occurrence and plays a major role in industry. However, these transfers can also become gateways for cyberattacks. The article explains how sensor technology can be protected against them.

When companies build IT infrastructures from numerous different manufacturers, the risk of incompatibilities grows. In this context, SAYTEC is positioning itself with its sayFUSE HCI system, which comes with IT infrastructure, a backup system and security technology built in.

From 18 October, NIS2 becomes mandatory in Germany: it tightens the EU-wide security requirements for operators of critical infrastructure (KRITIS). It classifies roughly twice as many sectors as critical, raises fines considerably and introduces new IT security obligations – for breaches of which senior management is liable.