A patient cannot check in. A customer cannot complete a payment. An employee cannot access the information needed to help either of them. Behind a service interruption sits a person whose plans have suddenly changed.
Incident management protects service delivery by coordinating how an organisation detects disruption, assesses its impact and restores service. The ITIL framework for IT service management places rapid restoration at its centre. As technology accelerates, that coordination deserves serious investment. [1]
AI is changing the pace of disruption
The UK’s National Cyber Security Centre (NCSC) assesses in its 2026 Annual Review that AI will increase the speed, scale and efficiency of cyber operations. It also warns of a widening gap between organisations able to strengthen their defences and those constrained by legacy technology, limited resources and weaker security programmes. [2]
Service disruption has many causes. The NCSC cautions that reported incidents do not follow a simple annual trend. Failed updates also demand a response: Microsoft estimated that the July 2024 CrowdStrike update affected 8.5 million Windows devices. [3] [4]
Google Cloud’s 2025 DORA research into software delivery adds another warning: AI tends to amplify an organisation’s existing strengths and weaknesses. My conclusion is straightforward: when delivery accelerates, testing, ownership and recovery capability must develop alongside it. [5]

Good process creates room for expertise
During a major incident, technical specialists need to investigate and restore systems while someone maintains the overall picture. Google’s incident response model separates incident command, technical operations and communication. Defined responsibilities help teams coordinate without making every engineer answer every stakeholder’s question. [6]
An incident process manager should make this possible before the emergency: agree priorities based on impact and urgency, maintain escalation routes, clarify supplier responsibilities, organise exercises and ensure the process improves. The incident commander then coordinates a specific major incident. Depending on the organisation, one person may perform both roles, but the responsibilities remain distinct. [1] [6] [7]
This is where organisational pace matters. A new tool may arrive before its support arrangements, recovery procedures and decision authority are clear. If teams must negotiate those responsibilities during an outage, recovery time is consumed by avoidable uncertainty. Leaders should fund preparation and exercises alongside the technology itself. [7]
Consider a payment outage. The response needs an agreed priority, someone authorised to approve a workaround, engineers investigating the fault and a named contact with the payment provider. A shared record should show what is known, what has been tried and who is doing what.
These arrangements give expertise somewhere useful to go. A room full of capable people still needs a shared objective.

Communication is part of service delivery
During disruption, people need information they can act on. Which service is affected? Is there a safe workaround? Should they wait, retry or use another channel? When will the next update arrive?
Google’s model gives a communications lead responsibility for regular updates and incoming questions. For customers and colleagues, a useful message should explain the confirmed impact, available alternatives and the next update time. If the recovery time is unknown, say so. [6]
For a hospital receptionist, that might mean knowing whether to continue with an approved manual check-in procedure. For a retailer, it might mean explaining which payment methods remain available. Communication helps people make decisions while technical recovery continues.
Crisis management connects recovery with leadership
Some incidents threaten essential operations, safety or the organisation’s ability to meet its commitments. Those consequences require leadership decisions alongside technical coordination: which activities should stop, which services take priority and which continuity arrangements should be activated?
The NCSC advises organisations to establish severity criteria, decision authority and communication responsibilities in advance, then practise their plans. Exercises should include relevant suppliers and service operators. [7]
Recovery must also be safe. In a cybersecurity incident, restoring access prematurely can leave a compromise unresolved. NIST’s 2025 guidance calls for preserving incident evidence, addressing the causes of compromise and verifying restored systems before returning them to production. Service availability, data integrity and security must be considered together. [8]
Before closing the incident, verify with service owners that people can complete the affected task again. A working login is only partial recovery if customers still cannot complete a payment. [8]

Every incident should improve the next response
After restoration, the organisation has another responsibility: act on what the incident revealed. Problem management investigates causes and helps reduce recurrence. A review should examine contributing conditions, including unclear ownership, fragile dependencies, poor alerts or an untested recovery procedure. [1] [9]
Google’s guidance emphasises learning without personal blame and assigning clear owners to corrective actions. Give improvements a priority, a deadline and a result that can be checked. An agreed action to test backups only becomes useful when someone performs the test and addresses what it reveals. [9]
I would track how long disruption lasts, how many people or critical activities it affects, how often it returns and whether agreed improvements are completed. A falling ticket count alone cannot show whether people are receiving a better service.
That is why incident and crisis management deserve investment in people, training and decision authority. Technology will keep changing. Reliable service delivery depends on an organisation that can respond coherently, recover safely and carry its learning forward.
Related reading: Root Cause Analysis in Incident Management: Beyond the Fix and IT Incident Management: More Than Fixing IT.
Sources and further reading
1. PeopleCert — ITIL 4 Practitioner: Incident Management. Official practice overview: restoration, roles, suppliers and problem management.
2. UK NCSC — Annual Review 2026: The impact of AI on cyber threat. Intelligence assessment of AI acceleration and an expected widening defensive divide.
3. UK NCSC — Annual Review 2026: Incident management. Explains why reported incident counts do not form a simple annual trend.
4. Microsoft — Helping our customers through the CrowdStrike outage (20 July 2024). Microsoft’s estimate of 8.5 million affected Windows devices, not organisations or incidents.
5. DORA — State of AI-assisted Software Development 2025. Research on AI as an amplifier of organisational strengths and weaknesses; not proof of universal causation.
6. Google — The Site Reliability Workbook: Incident Response. Practitioner guidance on command, operations, communication and preparation.
7. UK NCSC — Planning your response to cyber incidents. Board guidance on severity, delegated authority, responsibilities and exercises.
8. NIST — SP 800-61 Rev. 3 (April 2025). Cybersecurity incident-response guidance; evidence preservation and recovery validation (RS.AN-07 and RC.RP-05).
9. Google — The Site Reliability Workbook: Postmortem Culture. Practitioner guidance on blameless reviews, action ownership and follow-through.
Images are AI-generated editorial illustrations of fictional scenes.

