OneNym.idDiscuss deployment
English

THE BOARDROOM FIREWALL / Chapter 13

Incident Response and Crisis Communications

Every serious cyber incident tests leadership in the same ways: who decides, who is told, and what is said before the full picture is known. This chapter explains how to prepare for that day and how to lead through it calmly.

This page summarises the chapter. The full text is in the book.

The plan is the map; the drill is the habit

Incident response is the fire drill for cyber risk. Prevention is the sprinkler system; response is what people do when the alarm sounds anyway. An incident response plan is an agreement, made in advance, about who does what. It should define what counts as an incident and who can declare one, who leads and who stands in at night, which outside parties to call, who may speak to the press, customers and regulators, and which legal deadlines may start the moment the company becomes aware.

Two features are often missing. The first is an out-of-band channel, a way to coordinate if email and phones are compromised, such as a separate messaging group or a printed contact list. The second is a set of severity levels, so that a reported phishing click does not wake the board while a confirmed theft of customer data reaches it immediately. Above all, plans must be rehearsed. A tabletop exercise reveals gaps quickly, and every question left unanswered there is one the team will not face for the first time under real pressure.

Who decides what, and when

Unclear authority causes either paralysis or conflicting decisions. Well-run companies separate the technical response team, a named incident manager who coordinates and keeps the log, an executive crisis group that makes business decisions, and the board, which oversees and supports but does not run the response. A decision rights list settles in advance who can take a customer system offline, approve outside forensic help and its cost, authorise press statements, decide on customer notification, and approve any contact with attackers.

Containment often means switching things off, which costs money, so someone with business authority must decide quickly. Leaders should set a regular rhythm of updates in which the incident manager separates what is known, what is believed but unconfirmed, and what is still unknown. A decision log recording the time, decision-maker and reason for each significant choice becomes valuable evidence for regulators, insurers and auditors later.

Telling regulators, customers, staff and the market

Legal duties come first because they have deadlines. The book notes, for example, the general 72-hour rule for notifying regulators in the European Union and the United Kingdom, Japan's two-stage reporting for certain breaches, and a disclosure duty for listed US companies within a few business days of determining materiality. Many clocks start at awareness, not at the end of an investigation, so counsel should be involved from the first hours and most regimes accept an initial notice followed by updates. General counsel should own a current map of these deadlines before any incident.

Customers deserve a prompt, plain notice saying what happened, what data was involved, what the company is doing, what they should do and how to get help. Vague phrases about a sophisticated incident erode trust. Staff should hear early from leadership rather than from social media. Major customers often have contractual rights to be told within 24 to 72 hours. For the press, prepare a short holding statement, appoint one spokesperson, and never claim no customer data was affected unless that is certain.

Honesty, timing and outside experts

One principle guides every statement: say what you know, say what you do not know, and say when you will say more. Incidents tend to become public anyway, and companies that tell the truth early generally recover their reputation faster. Keep a steady rhythm of updates rather than a burst followed by weeks of silence.

Outside help should be chosen before it is needed: an incident-response or forensic firm, breach counsel, crisis-communications support and the cyber insurer, whose policy may require its own panel firms. Law enforcement can sometimes help, and in payment fraud speed matters because banks may be able to stop transfers if alerted within hours. Experts also distinguish two clocks: the technical clock of containment and the legal clock of notification. A recurring mistake is waiting for forensic certainty and missing a legal deadline.

The first 72 hours and the mistakes leaders make

The first hours are about confirmation and mobilisation, the rest of day one about containment and hard trade-offs, day two about scoping and first communications, and day three onward about sustained response, when fatigue becomes a real risk and rotations matter. In Meridian Foods' loyalty-database incident, the team filed with the regulator on what it knew and chose a plain-language customer email over an evasive one, and customers later said they appreciated being told directly.

Six leadership mistakes recur: hunting for blame too early, minimising the scale, going it alone without advisers, micromanaging the technical team, forgetting the exhausted people involved, and skipping the lessons-learned review. Leaders should also make clear that nobody engages with attackers except through the designated, professionally advised channel. The leader's main contribution is steadiness, because calm and panic both spread through a team.

KEY POINTS

Key points

  • Run a tabletop exercise with the executive team and track which gaps it revealed and which have been closed.
  • Set up an out-of-band way for the crisis group to communicate if email and phones are compromised.
  • Write a decision rights list covering system shutdowns, forensic spending, statements and customer notification.
  • Have legal own a current map of notification deadlines by country and sector.
  • Pre-select forensic, legal and crisis-communication advisers and check they fit your insurance policy.
  • Hold a blame-free lessons-learned review after every serious incident and track its actions to completion.

QUESTIONS

Common questions

What should a cyber incident response plan include?
A useful plan defines what counts as an incident and who can declare one, who leads the response and who deputises, which outside firms and authorities to call, who may speak to the press, customers and regulators, and which legal deadlines may apply. It should also include severity levels and an out-of-band way to communicate if email and phones are compromised. The plan only works if it has been rehearsed.
What is the board's role during a cyber incident?
The board's role is oversight and support, not running the response. It should ask whether the response is well led, whether the right outside help has been engaged and whether the company is meeting its legal duties. Directors who repeatedly contact the security team for updates, however well-meaning, pull people away from containing the incident.
When does the clock start for reporting a data breach?
Under many data-protection rules, the notification clock starts when the company becomes aware of a breach, not when its investigation is finished. That is why legal counsel should be involved from the first hours. Most regimes allow an initial notice based on what is known, followed by updates, and regulators generally prefer an early, incomplete notice to a late, polished one.
How should a company tell customers about a data breach?
Customers should be told promptly and in plain words what happened, what information was involved, what the company is doing, what they should do to protect themselves and how to contact the company. Vague phrases such as a sophisticated cyber incident tend to be read as evasion. People forgive accidents more readily than being kept in the dark.
What should we say to the press after a cyber attack?
Prepare a short holding statement that acknowledges the incident, says the company is investigating with expert help, and promises updates, and name a single briefed spokesperson. Avoid speculating about attackers, numbers or causes. Never state that no customer data was affected unless that is certain, because a retraction often does more damage than the breach itself.
What mistakes do executives make during a cyber crisis?
Common mistakes are looking for someone to blame too early, playing down the scale before it is known, deciding without advisers, micromanaging the technical team, neglecting exhausted staff and anxious customers, and skipping the lessons-learned review afterwards. Each comes from instincts that work in normal times but fail under crisis pressure.