When Systems Go Down: Building Resilience Before a Crisis

I recently joined Elliot Danker on Asia Pulse to discuss a question every critical infrastructure leader should be ready to answer: What happens when the systems you depend on fail? We discussed cyber resilience, secure communications, identity verification, cross-border cooperation, and crisis exercises. The central lesson was clear: organizations responsible for essential services must be able to operate when familiar technology is unavailable or untrustworthy.

Listen to the full conversation.

Key Takeaways for Listeners

• Plan on the assumption that critical systems will fail

• Maintain a secure, independent way to communicate during an outage

• Verify identities through stronger controls than voice or visual recognition

• Clarify ownership across operators, regulators, telecommunications providers, government, and technology teams

• Test realistic failure scenarios and use the results to improve

• Make every alert actionable and use responses to build situational awareness

Resilience Starts After Prevention Fails

Twenty years ago, most cybersecurity discussions focused on keeping attackers out. Prevention remains important, but it is no longer enough. Leaders must assume that an attacker may already be inside, a trusted system may become unavailable, or critical information may be inaccessible. Resilience means maintaining control, sustaining essential services, and making sound decisions after systems or networks have been compromised.

Resilience develops through planning and practice. Organizations should test how they would operate without email, identity systems, telecommunications, or operational networks. They must determine which services must continue, which decisions cannot wait, and who is authorized to act when approval chains fail.

Recovery is not the same as continuity. Restoring systems may take hours or days, while essential services must continue. A resilient organization can operate in a degraded state as technical teams contain the incident. Those priorities must be understood beyond the security team.

Communication Is the First Dependency

When a hospital, water plant, energy provider, or government agency loses its primary systems, the effects extend beyond IT. Teams may lose instructions, and leaders may be unable to confirm whether employees are safe. The organization still needs a trusted way to coordinate, issue instructions, and understand conditions.

Secure, out-of-band communications provide an independent channel. A backup that depends on the same identity provider, network, or infrastructure may fail with the primary system. Organizations should identify these dependencies and ensure authorized teams can quickly access backup communications.

Organizations also need control over communications. They should know where data is stored, who can access it, and who holds the encryption keys. These questions matter especially for governments and regulated industries, where sovereignty, jurisdiction, and accountability shape emergency requirements. Malaysia’s focus on sovereign, secure communications reflects the right question: Can a nation trust, govern, and control the systems on which its national security depends?

A communications system must remain available without weakening security. Teams also need a reliable way to verify who is giving an instruction. Synthetic voice and video can be convincing, so familiarity is no longer proof of identity. An urgent message that appears to come from a senior leader could redirect resources or interrupt operations. Strong technical checks, known escalation paths, and clear procedures must support trust. People need to know when to stop and verify.

Resilience Is a Shared Responsibility

Responsibility between connected systems is often unclear. I call this the “white-space gap.” Communications cross operators, telecommunications providers, regulators, government agencies, and technology teams. Each group may secure its environment while assuming another party owns the handoffs or emergency procedures.

Those gaps surface when several organizations must act at once. Who issues and verifies a warning? Who can change operations? Who communicates publicly? Clear ownership, named decision makers, and tested escalation paths reduce delay.

Cyber incidents cross organizational and national borders. Supply chains, cloud platforms, telecommunications networks, and shared infrastructure create dependencies no organization can manage alone.

Cooperation must begin before an emergency. Governments, infrastructure operators, regulators, universities, telecommunications providers, and security partners need shared playbooks, clear decision rights, dependable contacts, and realistic exercises. They also need enough trust to exchange relevant information quickly without exposing sensitive details. Disagreement must not prevent coordinated action when essential services and public safety are at risk.

Practice the Failure, Not the Ideal Response

A useful exercise removes the systems people expect to use. Shut down the network. Make alerting unavailable. Assume telecommunications have failed. Prevent key leaders from joining immediately. Then require teams to coordinate, switch to manual operations, and communicate under pressure.

Exercises should expose weaknesses while teams can still fix them. A failed notification, outdated contact list, unclear approval process, or missing manual procedure is a useful finding. Assign an owner, correct the problem, and test again. Exercises designed only to prove a plan works have little value.

Sending an alert is only the first step. People need clear instructions explaining what happened, what to do, and where to find verified updates. Vague warnings increase uncertainty.

Response teams also need field information: who is safe, who needs help, where operations have stopped, and where resources should go. These updates help build situational awareness. Two-way communication turns alerting into an operational tool and helps leaders adapt as conditions change.

The Question Every Organization Should Ask

Every organization should ask one blunt question: What would we do if everything went dark at 2 a.m. tomorrow?

And then: Who would be called first? How would leaders communicate? Which services would continue? Could teams verify instructions and operate manually?

If the answers are unclear, start with services that cannot fail. Map their technical and communications dependencies, assign decision owners, provide an independent way to coordinate, and test the plan.

The plan will evolve as systems, threats, and responsibilities change. What matters is starting before disruption forces the issue. When familiar systems are unavailable, preparation determines whether essential services continue.