Cybersecurity
The technology that protects everything else
Security products sit at the most privileged points of an organisation's infrastructure. That makes the question of who controls them, and under which law, especially relevant.
Why security technology is a special case
Most software processes data. Security software is trusted to see and control almost everything. To do its job, it typically needs:
- Deep privileges — endpoint agents at operating-system kernel level, identity systems that grant all access, PAM vaults holding administrator credentials, firewalls that can inspect and block all traffic.
- A continuous update channel — signatures, detection content and software updates pushed by the vendor, often automatically.
- Telemetry flowing outward — events, files and behavioural data sent to the vendor's cloud for analysis.
- Vendor access — support engineers, managed detection services and licensing back-ends.
This is why it is reasonable to ask not only how good a security product is, but also who ultimately controls it.
An illustration that is not about nationality
On 19 July 2024, a faulty content update for a widely used endpoint security product caused around 8.5 million Windows devices to crash, according to Microsoft's estimate. Airlines, hospitals and banks were affected worldwide.1
The vendor was not malicious and the issue was quickly identified. But the incident showed how a single vendor's update channel can affect critical operations across many countries at once. Control over that channel is a strategic dependency, wherever the vendor is based.
How dependent should European organisations be on non-European suppliers for the technology that protects their most critical systems?
European cybersecurity products are not automatically more secure. Security depends on engineering quality, secure development practices, vulnerability handling, transparency and track record. The argument here is narrower: geographic, legal, technological and economic dependency are real risk factors and should be included in risk assessments, alongside technical quality.
Layers of dependency in a security product
“European” or “non-European” is too simple. A product can be European at one layer and dependent at another. Map each layer separately.
Security technology areas and their sovereignty questions
Each area has its own dependency profile. Use the key question as a starting point for your evaluation, then check the European options.
| Technology | What it controls | Key sovereignty question |
|---|---|---|
| SIEM | Central log and event data of the entire organisation | Where are logs stored and analysed, and can the platform run fully under your control? |
| SOC technology | Detection, triage and response workflows | Can your SOC keep operating if a vendor cloud or managed service becomes unavailable? |
| EDR / XDR | Kernel-level agents on every endpoint and server | Who controls the update channel and the remote-response functions? |
| Firewalls | All traffic between network zones and the internet | Do updates, management and licensing depend on the vendor's cloud? |
| Network Detection & Response | Full visibility of network traffic | Is traffic metadata processed locally or sent to an external cloud? |
| OT security | Visibility and protection of industrial networks | Can it run fully on-premises in a segmented or air-gapped network? |
| ICS monitoring | Passive monitoring of industrial protocols | Can signatures and asset databases be updated offline? |
| PAM | Administrator credentials and privileged sessions | Who could, in theory, compel access to the vault or its key material? |
| Secure remote access | External access to internal and OT systems | Does every connection route through the vendor's infrastructure? |
| Identity & Access Management | Who can log in to everything | Can authentication continue if the identity cloud is unavailable or access is restricted? |
| Vulnerability management | Knowledge of all your weaknesses | Where is the list of your unpatched systems stored? |
| PKI | Trust anchors for devices, users and code | Who holds the root keys, and in which hardware? |
| Encryption | Confidentiality of data at rest and in transit | Who controls the keys — you, or the service provider? |
| Email security | All inbound and outbound mail | Where is mail content processed and quarantined? |
| Threat intelligence | Situational awareness and attribution | Whose perspective and priorities shape the intelligence you receive? |
| Secure collaboration | Confidential communication and documents | Is end-to-end encryption under your control, and where are servers operated? |
| Cloud security | Configuration and workloads in the cloud | Does the tool add a second dependency on top of the cloud provider? |
| Backup | Your last line of defence against ransomware | Can you restore without the vendor's cloud, licence server or support? |
| Incident response technology | Forensics and containment during a crisis | Are tools and retainers available when you need them, under suitable jurisdiction? |
Four dependencies to add to your risk assessment
Risk frameworks already cover confidentiality, integrity and availability. Supplier dependency is often handled only through contracts. Make it explicit.
Geographic
Where are the people, data centres, update servers and support teams? Which of them sit outside Europe?
Legal
Which jurisdictions can compel the supplier to disclose data, suspend service or restrict exports?
Technological
Can you operate, update and restore without the supplier's cloud? Are formats and protocols open?
Economic
What happens if the supplier is acquired, changes its pricing, ends a product line or becomes insolvent?
Under the NIS2 Directive, essential and important entities must already address supply-chain security, including the security-related aspects of relationships with direct suppliers.2 Sovereignty considerations fit naturally into that existing obligation — but NIS2 does not require buying European. See regulation vs. choice.
OT & critical infrastructure
When technology dependency becomes a societal risk
In energy, water, telecom and transport, a supplier problem is not only a business problem. It can affect public safety and the functioning of society.
Energy & electricity
Grid control, substations, generation and distributed energy resources.
Gas
Transmission, compression, storage and metering systems.
Water
Drinking-water treatment, wastewater and flood control.
Telecom
Core networks, radio access and network management.
Transport
Rail signalling, traffic management, ports and aviation.
Government
Public administration, emergency services and defence-adjacent systems.
Industry
Manufacturing, chemicals and process industries.
OT, ICS & SCADA
The control systems that move physical processes — often in service for decades.
Why sovereignty is especially relevant here
- Long lifecycles. OT systems often stay in service for 15 years or longer — far longer than the typical horizon of a vendor's strategy, ownership or pricing.
- Physical consequences. Availability and safety come first. Losing access to updates or support can have consequences in the physical world.
- Segmented networks. Many OT networks are, or should be, isolated. Products that require continuous cloud connectivity can conflict with that design.
- Regulation. Operators in these sectors fall under NIS2 and, for designated critical entities, the Critical Entities Resilience (CER) Directive, which focus on resilience and continuity.23
- Geopolitics. Export controls, sanctions and trade disputes can affect the availability of technology and updates — a scenario that should be planned for, however unlikely it seems today.
Stress-test scenarios
Run these scenarios against each critical supplier:
- The vendor is acquired by a company from another jurisdiction.
- A government restricts exports of the technology or its updates.
- The vendor discontinues its on-premises version in favour of SaaS.
- The vendor's cloud service is unavailable for a week.
- The vendor's update channel is compromised.
- The vendor becomes insolvent.
Twelve questions for OT and critical infrastructure suppliers
- Can the system operate without an external cloud service?
- Can security updates still be obtained during a geopolitical disruption?
- Who controls the update infrastructure?
- Where is the source code maintained?
- Can the product be operated fully on-premises?
- Does licensing require continuous internet connectivity?
- Where is telemetry processed?
- Can the organisation export all configuration and security data?
- Is there an alternative supplier?
- What happens if the vendor is acquired?
- What happens if a foreign government restricts technology exports?
- Which jurisdiction ultimately applies to the supplier?
Prefer products that can run fully on-premises and offline; mirror updates internally and validate them before deployment; negotiate source-code escrow and long-term support commitments; require open, documented protocols and export formats; keep a qualified second supplier for the most critical functions; and follow the IEC 62443 series for industrial security.