<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=2171572209666742&amp;ev=PageView&amp;noscript=1">
men in a business suit sitting at a desk behind a laptop comuter

Digital resilience: staying operational when it matters most

Six risk questions to set clear priorities

 

Cyberattacks, system outages, AI, and growing dependencies on technology providers are constantly changing the risk landscape. Responding with an ever-growing set of standalone measures leads to a loss of focus. What matters is the big picture: What poses the greatest threat to critical services, how are the risks interconnected, and what needs to be addressed first?


Six risk questions help companies and public authorities build this understanding and set clear priorities.

icon

Key takeaways

  • Digital resilience means staying operational when things go wrong.

  • Many organizations understand their individual risks but lack the big picture.

  • Adnovum uses six risk questions to develop a comprehensive assessment, including the key risks.

What is digital resilience?

Digital resilience is an organization’s ability to maintain critical business and public administration functions during cyberattacks, system outages or provider disruptions – or restore them within a defined timeframe – and demonstrate this capability to regulators, auditors, and customers.

 

A resilient organization understands its dependencies across data, infrastructure, and applications, has tested its business continuity scenarios and knows which risks to mitigate first.

Why digital resilience is a leadership responsibility


The cost of an outage rarely stops at IT. When customers cannot make payments, production comes to a halt, or public services become unavailable, a technical issue becomes a business issue – with consequences for reputation, compliance, and trust.

Digital resilience therefore starts with a leadership question: which outages are acceptable, and can we provide evidence for that? Backup strategies, access rights, provider contracts, training, and AI policies all demand attention. CEOs, CIOs, CISOs and public sector leaders must turn many legitimate concerns into clear priorities. To do so, they need to know where their organization is most vulnerable. Often, the honest answer is: we have an idea. A rough one.


Roughly isn't good enough. Regulators, customers, and contractual partners increasingly expect evidence, not assurances. Staying operational means understanding and mitigating risks: having the big picture, setting clear priorities, and taking effective action.

Commonly confused: digital resilience, cyber resilience, and digital sovereignty

Cyber resilience is an organization’s ability to stay operational during cyberattacks. Digital sovereignty describes control over data, technologies, and providers. Digital resilience encompasses both, connecting them with service continuity, governance, workforce enablement, and the management of AI.

In short: sovereignty is a means; resilience is the business objective.

Learn more about Cyber Resilience

Plenty of pieces, no big picture 

 

Organizations often lack an understanding of the interdependencies among risks and which ones carry the greatest weight. They usually have no shortage of measures. IT security protects access. Risk management uses the business impact analysis (BIA) to establish which outages are acceptable. Compliance prepares for the next audit. Without a shared assessment, priorities are driven by whatever creates the most pressure: the latest incident, the next audit, or the loudest voice in the room. Meanwhile, the greatest risk may go unaddressed.

 

Security assessments, business continuity management (BCM) analyses, and sovereignty reviews each deliver valid findings, but usually only part of the picture. The connections between risk areas reveal where decisions are needed. A second data center offers little protection if both sites depend on the same provider. Robust access management still leaves gaps if no one has defined what an AI agent is allowed to do.

 

Adnovum’s six risk questions bring these pieces together. They provide the big picture of the risk landscape, enabling organizations to prioritize risks by business impact and translate findings into actionable measures.

Understanding risks: six answers, three outcomes

Digital resilience is reflected in three outcomes: decisions are accounted for, access is clearly defined, and operations keep running. Each outcome is supported by two of Adnovum’s six risk questions. Together, they reveal whether an organization can still act in a crisis. Resilience is an ongoing process: organizations regularly check whether their measures are effective and whether new dependencies have emerged.

Outcome 1: decisions are accounted for

Decisions (governance) and data access (data control and sovereignty) can be traced and substantiated for third parties.

Colleagues reviewing a document.

Who made the decision, why, based on what authority?

Governance 


Good governance is reflected in the answer to one question: why was this decision made? Decisions that cannot be explained create compliance and reputational risks.

AI raises the stakes. When a system makes a lending decision or blocks a transaction, the organization must be able to explain why, who made the decision and who is accountable. This is where AI governance comes in: rules and controls that define what AI is used for and who is responsible for its outcomes. The EU AI Act also affects Swiss organizations when their AI systems are made available in the EU or their outputs are used there.

Adnovum builds AI governance frameworks with controls for traceable AI decisions, establishes information security management systems (ISMS) and secures privileged access with a comprehensive audit trail.

Example Versicherung Schweizer Ärzte

Server racks in a modern office.

Who has legal access to the data, where is it stored, and who is accountable?

Data control and data sovereignty

Data sovereignty means verifiable control over access, storage location, encryption keys, and portability.

Organizations typically lose control of their data because the legal basis, access responsibilities, and hosting responsibilities are unclear – for example, when personal data is held in third-party solutions or customer data is stored in clouds outside Switzerland and the EU.

Since September 2023, the revised Swiss Federal Act on Data Protection (FADP) has imposed stricter requirements for demonstrable compliance and data protection management. Where there is an EU connection, the GDPR also applies. The Swiss federal government and cantons are also expected to make data available in an interoperable format without losing control over it.⁴ Swiss hosting alone does not meet these requirements. Clear responsibilities and verifiable control over critical data and systems are essential.

Adnovum integrates the Swiss e-ID as a basis for verifiable data access, supports data protection management and compliance with the revised FADP, publishes public sector data in a controlled manner as Linked Data, and provides hosting under Swiss law.

Example Viscom Engineering

Outcome 2: access is clearly defined

Every access can be linked to an identity (identity security), and every action is governed by a clear framework (enablement and control).

Woman using a phone in an office.

Is every access authorized and verifiable, whether by a person or an AI agent? 

Identity and cyber defence 

Under the Zero Trust principle, no access is trusted solely because of where it originates – whether from people, machines, or AI agents. Identity, device, context, and authorization are continuously verified, and access rights are limited to the minimum necessary. This puts identity security at the heart of cyber defence.

Attacks increasingly target identities through stolen credentials, phishing, and misused permissions. AI agents introduce additional non-human identities into systems that traditional perimeter security does not cover. The goal is visibility across all access: every access can be linked to an identity, systems are continuously monitored, and prevention, detection, and response work together. Since 1 April 2025, operators of critical infrastructure have been required to report cyberattacks to the Swiss National Cyber Security Centre (NCSC) within 24 hours. The EU’s NIS2 Directive does not apply directly in Switzerland, but has an impact through contracts and procurement with an EU connection. The challenge is usually prioritization: which access points need the strongest protection?

Identity and access management (IAM) has been a core area of Adnovum’s expertise since the 1990s: for customers (CIAM), employees (EIAM), and privileged accounts (PAM). Adnovum integrates the Swiss e-ID, assesses systems through penetration testing, and secures AI systems. From its own Security Operations Centre (SOC) in Switzerland, certified to ISO 27001 and PCI DSS Level 1, Adnovum monitors systems 24/7, detects attacks early, and contains them through incident response.

Examples SIX und Cornèr Banca

Colleagues discussing in a meeting.

Do employees and AI agents know what they are allowed to do?

Enablement and control

People need to be enabled to act; AI agents need clear boundaries. Clearly defined limits must apply to both.

Employees who fail to recognize threats leave the door open to attackers. At the same time, AI agents are increasingly acting on behalf of organizations. The goal is for the organization to know at all times who performed each action – a person or an agent. Only then can it explain why an action was taken. Financial institutions subject to FINMA Circular 2023/1 must also train their employees and crisis management organization in BCM, with training tailored to their size and risk profile.

Adnovum trains employees of client organizations in information security (security awareness), runs phishing simulations, and defines permissions and operating boundaries for people and AI agents.

Outcome 3: operations keep running

Critical functions keep running even if a system or provider fails (service continuity and ability to switch providers).

Computer displaying a progress bar.

Can operations continue if systems fail?

Service continuity and recovery

Service continuity means restoring critical functions within a defined timeframe.

Many organizations are more centralized than they realize: a single data center, a single critical provider. If a ransomware or DDoS attack hits one of these points, operations stop. Service continuity therefore starts with critical services: core public services for authorities, or production, payments and customer service for businesses. Only when the maximum tolerable downtime for each function is clear can an organization decide where redundancy is needed. BCM, incident response, and disaster recovery must be documented and rehearsed. Financial institutions subject to FINMA Circular 2023/1 define disruption tolerances for this purpose. They test these regularly, including against severe but plausible scenarios, with the scope tailored to their size and risk profile. FINMA refers to this as operational resilience.

Adnovum operates systems with clear service level agreements (SLAs), redundancy and recovery strategies, and designs distributed infrastructures. The SLAs define in advance who does what and by when – including in the event of a cyberattack.

Business handshake across a desk.

Can providers be switched without disruption?

Ability to switch and provider independence

Dependencies on individual providers can rarely be avoided altogether. But they can be managed in ways that keep switching options open.

Dependencies often develop unnoticed over the years through proprietary platforms, licensing models, and a lack of interoperability – commonly known as vendor lock-in. As regulation increases, provider concentration is becoming a focus of scrutiny.

The goal is exit readiness: the ability to switch providers without interrupting operations because portability is built into the architecture. Unlike complete independence, exit readiness can be tested and backed up with evidence. This makes it the aspect of digital sovereignty that can be proven in day-to-day operations. FINMA identifies the insolvency of a key service provider as a scenario for resilience testing by supervised financial institutions.

Where there is an EU connection, the EU’s Digital Operational Resilience Act (DORA) requires exit strategies for critical ICT service providers.

Adnovum uses open standards and open source, develops vendor-neutral cloud strategies with portable workloads and makes supply chain dependencies assessable and manageable.

Setting priorities: from assessment to decision

 

The six risk questions reveal where an organization is vulnerable. The next step is deciding what to address first. No organization can eliminate all risk. The goal is to make informed, evidence-backed decisions about which risks are acceptable and which must be mitigated. This is how organizations stay operational when it matters most.

 

Adnovum works with the organization to assess which outages are tolerable, which obligations apply and which dependencies connect multiple risks. The result is a substantiated action plan with clear responsibilities, investment decisions, and implementation plans that the organization can justify to its board, auditors, and regulators. Priorities are transparent and tailored to the organization: what is secondary for a business may be a priority for a public authority. For the Zurich Cantonal Tax Office, Adnovum assessed, quantified, and prioritized IT risks and developed 50 risk mitigation measures.

 

Resilience is an investment in prevention. Its benefits can be difficult to quantify. This makes an evidence-based business case all the more important, drawing on the BIA, quantified risks or, regulatory obligations.

Portrait of Beat Fluri, CTO at Adnovum

Act effectively: four steps into operation

Adnovum supports organizations through a four-step cycle from assessment to operations: assess the risk landscape, set priorities, implement measures, and operate and demonstrate resilience.

You gain an evidence-based view of your risks, a substantiated action plan and measures that deliver results in operation.

Assess the risk landscape: We work through all six risk questions with you to assess your risk landscape. We build on the foundations you already have, e.g., BIA, ISMS, BCM documentation, data protection inventories, architecture overviews, supplier assessments, and audit reports.

Set priorities: We assess each risk based on business impact, regulatory obligations, and dependencies. We translate the assessment into a substantiated action plan that you can justify to your supervisory bodies.

Implement measures: We implement the highest-priority measures, taking a vendor-neutral approach and drawing on our own development and security teams.

Operate and demonstrate resilience: We operate the solutions under clear SLAs and rehearse crisis scenarios with you. This means your resilience is tested and can be demonstrated to third parties. We then reassess the risk landscape.

What sets Adnovum apart

Adnovum is a Swiss IT service provider for companies and public authorities. Since 1988, we have secured critical systems for the financial sector as well as the federal government and cantons. Our recommendations are vendor-neutral: we spun off our own Nevis security platform in 2020. And because we implement and operate what is set out in the concept, we set realistic priorities. Guided by these principles, our client relationships last an average of 10 to 15 years.

Decide today, act decisively in a crisis


Priorities, roles, and alternatives must be clear before a crisis strikes. Organizations that manage digital resilience as an ongoing process strengthen their ability to act and can demonstrate to governing bodies, regulators, customers, and the public why particular measures take priority.

The key question is therefore: do we know which risks we can accept and which we need to mitigate – and can we substantiate that? Organizations that can answer this no longer have just a rough idea of their vulnerabilities. They know where to start.

Adnovum works with you to analyze risks, set clear priorities, and implement measures through to operations – so you can stay operational.

Five questions for your next board or executive management meeting

 

Use these questions to assess the state of digital resilience in your organization:

1. What does digital resilience mean for our organization?

2. Which outages can we accept, and what can we not?

3. Can we substantiate our priorities to auditors and regulators?

4. What do FINMA, DORA and NIS2 require of us?

5. How quickly can we restore operations after an outage?