25.02.2026 / 08:30
Ten Steps for Dealing with Cyber Risks
Energy Markets

Cyber risks are among the greatest threats today, yet they are also the least covered. Few decision makers understand cyber security issues. Measures are not always implemented consistently enough.

Cyber security issues include defending against threats such as malware and ransomware, securing networks, access, and data, as well as strengthening risk management, contingency planning, and employee training.
Most companies must also deal with security for AI, the cloud, and the supply chain. In addition, security systems must be checked regularly. Bringing in an organization that specializes in cyber defense ensures Confidentiality, Integrity, and Availability (CIA).
The CIA Triad
To maintain business operations without disruption, the CIA triad is critical.
Integrity is best covered by the triad, which happens primarily in applications. Digitalization and automation ensure more consistent application coverage of processes. This leads to increased integrity as a by-product, through better auditability via logs and fewer process interruptions.
Availability is threatened by ransomware encryption of systems and distributed denial-of-service (DDoS) attacks. While these can usually be resolved by a specialized network provider, the ransomware problem runs deeper. Confidentiality is under attack on several fronts. Ransomware attacks are increasingly focused on stealing data, rather than just encrypting systems. To prevent disclosure, money is extorted and data is offered for sale.
Another potential threat is employees stealing data for later use. The risk of illegal access by governments has also increased. More must be done to protect data in connection with AI.
Step 1: Develop ISMS
An information security management system (ISMS) creates clarity, prevents important risks from being overlooked, and protects against unnecessary investments.
Availability
Protect your resources from external manipulation, manage necessary capacities and ensure that errors are detected preventively or as quickly as possible.
Step 2: Protect Systems
Antivirus software, firewalls, and strong authentication are essential to prevent attackers from taking over systems. “Living off the land” attacks exploit existing resources. Access to these resources should be prevented wherever possible.
Step 3: Prevent DDoS
Engage your network operator to prevent denial-ofservice attacks. This must be done by a specialist with sufficient resources.
Step 4: No Reaction without Detection
Good monitoring of all resources is important, as attacks cannot be prevented 100% of the time.
Step 5: Deploy Rapidly
Deploying new resources must happen quickly on clean sources, so a secure environment can be swiftly restored. Kubernetes is one of the most common methods. “Infrastructure as code”, in conjunction with hyperscalers, also offers a good measure of high availability.
Confidentiality
Confidentiality is based on two principles: data can be locked or encrypted, while all other measures essentially relay in trusting the providers. However, in a zero-trust approach, that implicit trust in data storage and management is deliberately removed.
Step 6: Use IAM
Authorization of access (IAM) is a mandatory requirement for people, machines, services, and AI agents.
Step 7: Draw a Perimeter with DLP
Data should be where it is needed. Everywhere else, it is unnecessary. If the environment is under full control, establishing a perimeter, through data loss prevention (DLP) or resource separation, is sufficient. Since DLP is required, it is advisable to start with it first. However, segregation impacts the entire architecture, which makes it costly to install.
Step 8: Encrypt
If a system cannot be fully controlled, it poses a challenge to sovereignty. This can be addressed by using more local resources, through greater trust in contracts, or with encryption.
More local resources require established tools, capacities, and functionalities. Europe has lagged behind in this area. There is even a risk that AI increases dependencies.
Trust through better contracts might be difficult in the current geopolitical situations. Hence, encryption remains the only feasible and sustainable solution to address the sovereignty.
Step 9: RMS and CDPG
Data encryption depends on representation of the data. Files are encrypted using rights management solutions (RMS). Data in structured form requires fieldlevel encryption of tokenization, usually provided by cloud data protection gateways (CDPG).
Cyber Security in Municipal Utilities
The resources available to municipal utilities are limited. Most expenditure is therefore absorbed by day‑to‑day operations. Cyber security should be integrated into day-to-day operations on a risk-based basis. For the measures to be effective, they must not be treated as a minor issue.
Step 10: Keep It Simple
An integrated risk management system is necessary to identify critical aspects and address them directly. Avoiding unnecessary actions is as important as taking the appropriate measures. This is the only way to make optimal use of available resources.
Cyber Security in Energy
Energy must flow. Availability is therefore the most critical factor. As a highly attractive target, defense against external attacks is of primary importance. Therefore, no compromises should be made on steps 1 to 5.
A network of suppliers makes integrity a significant issue. The chain breaks at its weakest link. For integrity, this means that no link can be weaker than the entire chain. This requires standards, regulations, and a mutual understanding.
AI
The data hunger of AI is insatiable. Without data for training, the models will not improve. Exposing data only where necessary is an important factor.
It will be difficult to operate without AI in the future. That is why two other aspects are currently becoming more important:
1. Prompt management prevents sensitive information from reaching AI via queries (prompting).
2. Agentic AI is the new hot topic. Countless agents with specialized tasks collaborate to achieve an overall result. For this to work, these agents often have extensive permissions. To ensure they do not undermine the “least privilege” effort, every action taken by these agents should be analyzed on a case-by-case basis.
Next Steps
This article has outlined the overarching direction and key considerations for addressing these challenges. It is not intended to define a fully integrated risk management framework or a comprehensive ISMS. Where these structures are not yet in place, establishing them should be the immediate next step and the foundation for future progress.
Ransomware Considerations
Preventive measures include backups, awareness training for employees, and keeping a watchful eye. With AI-supported attacks, their quality typically increases faster than employee awareness.
To avoid blackmail over content, and to prevent it from being sold to third parties, encryption is recommended. This removes a significant part of the attackers’ business case, causing them to focus on easier targets.





