Trust Requires More than Just Documentation
Security Assurance Reimagined
KEY FACTS AT A GLANCE
Digital supply chains extend a company’s risk zone far beyond its own boundaries. This also changes the nature of trust: Customers increasingly expect solid evidence that security controls are not only defined but actually work.
How Much of Your Security Architecture Do You Actually Control Yourself?
Cloud platforms, software vendors, managed services, and specialized service providers are now an integral part of nearly every IT landscape. As a result, an organization’s security architecture no longer ends at the corporate perimeter. The Verizon Data Breach Investigations Report 2026 highlights the significance of these dependencies: Third parties were involved in 48 percent of the breaches investigated. (https://www.verizon.com/business/resources/reports/2026-dbir-data-breach-investigations-report.pdf)
This presents two opportunities for companies:
- As a customer, I need to understand the risks a supplier introduces into my architecture.
- As a service provider, I am increasingly required to demonstrate that my own controls are effective.
Trust alone is no longer enough.
Third-Party Risk Is Not a One-Way Street
Those who outsource a service can delegate tasks. However, the associated risk does not automatically disappear. That is why NIS2 makes supply chain security an integral part of cyber risk management. When evaluating suppliers, factors such as their security practices, secure development processes, and supplier-specific vulnerabilities play a role. (https://eur-lex.europa.eu/eli/dir/2022/2555/oj)
The ENISA NIS Investments Report 2025 shows that this trend has taken hold in practice: 90 percent of the organizations surveyed from high-criticality NIS2 sectors are implementing specific measures for third-party and supply chain security. 63 percent require security standards or certifications, 54 percent conduct supplier risk assessments or audits, and 48 percent incorporate cybersecurity requirements into contracts. (https://www.enisa.europa.eu/publications/nis-investments-2025)
The crucial question, then, is not: Can I trust my service provider? But rather: How can this trust be reliably justified?
Documentation Is Important. But It Is Not Evidence.
The standard response to this often involves documentation: policies, questionnaires, certifications, and audit reports. This is correct and important. An ISO certification or an independent audit report provides external validation and reduces the need to thoroughly review every control system on your own.
It becomes problematic when this is treated as the actual proof of security. A policy describes what should happen. A questionnaire is, at first, simply a structured self-assessment. And a certificate does not automatically confirm that every relevant control for a specific service is functioning effectively at all times.
The requirements for managing third-party ICT service providers under DORA make this distinction clear. Financial institutions may use certifications and audit reports to monitor their ICT service providers. However, they should not rely exclusively on these in the long term. Among other things, the scope and timeliness must be assessed, and the operational effectiveness of key controls must be taken into account.
(https://eur-lex.europa.eu/eli/reg_del/2024/1773/oj)
Certification builds trust in a control system. Operational evidence shows that this control system actually works. Both have their place.
Evidence Doesn’t Start with Technology
Evidence by Design doesn't mean simply collecting more telemetry or building a dashboard for every control. The prerequisite arises earlier.
Organization is the craft; technology is the tool.
Governance establishes what is expected. Business processes translate these expectations into an actionable workflow. Security defines the necessary controls. Only then can technology help to reliably enforce these controls and provide verifiable evidence of their implementation.
For me, this can be broken down into a simple chain:
Governance > Process > Control > Technical Implementation > Evidence
This means that evidence doesn't become an additional step in preparation for the next audit. It is generated right where the actual work takes place.
Audits Generate Operational Evidence
Here’s an example from software development: A secure development policy can require reviews, security tests, and approvals before a production deployment. This process can be documented and audited later. Alternatively, the controls can be built directly into the development process. The DevOps system then documents changes, reviews, approvals, and security tests. The pipeline can even prevent deployments from taking place without the necessary controls.
The ENISA NIS2 Technical Implementation Guidance also describes exactly this type of evidence. For secure development, it lists, among other things, version control, change management logs, code reviews, project management systems, and current test results. For identity management, for example, it mentions review and approval logs. (https://www.enisa.europa.eu/publications/nis2-technical-implementation-guidance)
These examples do not constitute binding law. However, they provide concrete examples of how regulatory requirements can be translated into verifiable operational practices.
"Evidence by Design" means that controls are integrated into business and technology processes in such a way that their execution itself generates robust evidence.
The same principle applies to identity management, change management, vulnerability management, backup and recovery, and business continuity. Automation is extremely helpful in this regard, but it is not a prerequisite.
Regulation Moves Through the Supply Chain
For service providers, this capability is becoming economically significant. The NIS2 Implementing Regulation (EU) 2024/2690 illustrates this for its specific scope of application. It covers, among others, cloud and data center providers, managed service providers, and managed security service providers. For the entities covered, the regulation also specifies requirements for the management of suppliers and service providers. Contractual agreements should, where appropriate, cover, among other things, cybersecurity requirements, incident reporting obligations, the handling of vulnerabilities, requirements for subcontractors, and audit and verification rights.
(https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32024R2690)
When a regulated company needs to assess its supplier risks, it requires evidence from the service provider. The service provider, in turn, may need evidence from its own subcontractors.
Regulatory requirements thus flow through the supply chain via contractual and procurement relationships.
Evidence Readiness Becomes a Service Capability
This makes assurance an integral part of the service. Audits, certifications, penetration tests, technical documentation, supplier governance, and customer-specific assessments all require resources. If the same evidence has to be gathered anew for each customer from ticket systems, DevOps platforms, IAM solutions, and documentation, compliance does not scale well.
Evidence by Design changes this cost logic:
The first verification requires architectural effort. The hundredth should require as little additional effort as possible.
An Internet service provider should therefore consider accountability to be just as much a part of its service as availability, support, or performance.
Evidence Readiness thus becomes a service capability.
Organizations that can generate relevant evidence from their services and processes in a structured manner can complete due diligence processes more quickly, meet recurring requirements more efficiently, and demonstrate to customers in a more transparent way how security is actually implemented. It is explicitly not a matter of giving customers full access to internal security systems. Evidence must be provided in a risk-based manner, appropriately abstracted, and controlled with regard to its confidentiality.
From Evidence to Continuous Assurance
The long-term goal, therefore, is not simply to create a larger evidence repository. If controls are already integrated into operational processes, deviations can increasingly be detected during operations. A missed review, an unauthorized change, or an overdue recertification then does not have to be discovered months later during an audit. In this way, evidence becomes the foundation for active management.
Compliance without oversight is mere self-reassurance.
Oversight without evidence remains mere assertion.
Evidence by Design combines both with operational reality. The practical question, therefore, is:
Which of your critical controls already generate reliable evidence from day-to-day operations, and for which ones does the evidence still need to be manually reconstructed for customers, audits, or regulatory reviews?
Written by
Kai Korla works at Arvato Systems at the intersection of cybersecurity, governance, and architecture. He focuses in particular on how security risks can be not only documented but also actively managed through clear accountability and informed decision-making.