On this page
It is 06:40 on a Monday. The finance manager rings to say every file on the shared drive now ends in a strange extension, and a text file called README_RESTORE sits in every folder. In the next hour, the people in your business will make decisions that decide whether this is a bad week or a bad year. Most of those decisions are not technical. They are about who is in charge, what gets switched off, what gets kept, and who gets told.
This is the runbook we work from for the first 24 hours of a ransomware incident, written for the people in the room: the managing director, the IT lead, operations, and whoever handles legal and communications. It follows guidance from CISA, the UK NCSC and NIST. If your question is still "is this an outage or an attack?", start with how to tell an outage from a cyberattack and come back here once the answer is "attack".
Why the first day decides the outcome
Ransomware is rarely the start of an attack. It is the end. By the time files are encrypted, the attacker has usually been inside for a while, gained admin access, looked for backups and often copied data out. Encryption is where they stop hiding.
So you are not only fighting encrypted files. You are dealing with someone who may still be inside, still hold admin passwords and already have a copy of your data. A plan that only asks "how fast can we restore?" restores you straight back into their hands.
CISA's #StopRansomware Guide splits the response into detection and analysis, containment and eradication, then recovery. NIST SP 800-61 Revision 3 treats incident response as part of risk management, not a purely technical task. The runbook below follows that order, mapped to the clock.
Hour 0 to 1: take control and stop the spread
The first hour is about two things: putting one person in charge and stopping the damage from spreading.
Name an incident lead. One person makes the calls. In a small company this is often the COO, not the most technical person, who is needed hands-on. Name a deputy too.
Move off your own systems for communication. If the attacker holds domain admin rights, they may be able to read your email and chat. Agree a fallback before you need it: a group on personal phones, a separate messaging account, or a conference bridge that does not sign in through your company directory. Do not discuss the response, ransom or suspects on corporate email until you know it is clean.
Isolate, do not power off. CISA's guidance is to identify the affected systems and isolate them immediately, and to power devices down only if you cannot disconnect them from the network. The reason is evidence. A running machine holds things in memory, such as running processes, network connections and sometimes encryption keys, that vanish when it is switched off. In practice isolation means:
- pulling network cables or disabling Wi-Fi on affected machines;
- blocking affected hosts at the switch or firewall;
- using your endpoint protection tool's "network contain" function, if you have one, which cuts the machine off from everything except the security console.
Protect your backups. If backup servers or storage are still reachable from the network, disconnect them now. Attackers target backups, and may try again once they notice the response.
Start the log. Write down every action with the time, who did it and why. Your insurer, lawyers and possibly a regulator will read it later.
# Incident IR-2026-001 lead: A. Example deputy: B. Example channel: out-of-band group06:52 [A] Lead appointed. Out-of-band group created. Corporate email NOT used for IR.06:58 [B] FS01 and 3 finance laptops isolated via EDR network containment. Left powered on.07:05 [B] Backup appliance 10.0.0.40 unplugged from network. Not powered off.07:10 [A] Insurer hotline called, claim ref pending. Told: do not contact the attacker.07:20 [!] Ransom note found on FS01. Photographed. Not opened on any other machine.Hour 1 to 4: understand the scope
With the spread slowed, the next question is how far the attacker got. Resist the urge to start fixing machines. You need a picture first.
Make the calls you are obliged to make. If you have cyber insurance, ring the insurer's incident line early. Many policies require prompt notice and expect you to use their approved response firm and lawyers; acting first and telling them later can put cover at risk.
Preserve evidence before you change anything. CISA's checklist calls for a system image and a memory capture from a sample of affected devices. Preserve logs too: firewall, VPN, identity provider, email and endpoint protection logs. Many cloud and appliance logs are only kept for a short time by default, so export them now. In our engagements, the logs that matter most are usually the ones nobody thought to keep: VPN sign-ins and the admin activity on domain controllers.
Answer four scoping questions. Write the answers in the log, marked as confirmed or suspected:
- Which systems are encrypted or affected? Servers, laptops, cloud storage, backups.
- Which accounts did the attacker use? Look for new admin accounts, unusual admin sign-ins, and service accounts acting at odd hours.
- How did they get in? Common routes are a VPN or remote access without MFA, an unpatched internet-facing system, stolen credentials or a phishing email.
- Was data taken? Look for large outbound transfers, archive tools such as 7-Zip or WinRAR on servers that do not normally run them, and cloud sync or file-transfer tools nobody installed.
Identify the variant, carefully. The ransom note and file extension often name the ransomware family, which tells you how the group usually behaves and whether a free decryptor exists (the No More Ransom project lists known ones). Do not upload files containing customer data to public analysis sites.
Hour 4 to 12: contain the attacker, not just the malware
By now you should know roughly where the attacker went. Containment means taking away their access, not only stopping the encryption.
Reset credentials in the right order. Start with the accounts the attacker is most likely to hold: domain admins, other privileged accounts, service accounts and the accounts used for backup and remote management. In Active Directory, the krbtgt account's password must be reset twice, with time between the resets for replication, to invalidate Kerberos tickets the attacker may have forged. Plan this with your response team, because resetting it badly can break authentication across the domain. Our post on Kerberos delegation in Active Directory explains why Kerberos settings matter so much after a compromise.
Revoke sessions and tokens in the cloud. Resetting a password does not always end a session the attacker already holds. In Microsoft 365, revoke the user's sessions as well. Check for new mailbox rules, OAuth app consents and MFA methods added during the incident. We cover the sign-in side in detecting stolen-session sign-ins in Entra ID.
Close the entry point. If the attacker came in through a VPN without MFA, turn on MFA or turn the VPN off. If it was an unpatched system, take it offline or patch it. Recovery that leaves the door open usually ends with a second encryption.
Kill the tooling. Block the ransomware binaries you have found in your endpoint tool, and look for remote access tools the attacker left behind.
Hour 12 to 24: decide, communicate, plan recovery
The second half of the day moves from firefighting to decisions. This is where the leadership team earns its keep.
Decide what to restore first. List the business services in the order the company needs them, not the order IT finds easiest. Payroll, customer-facing systems and the tools staff need to work usually come first. Restore from backups you have checked are clean, onto rebuilt or verified systems, on a network segment the attacker cannot reach. CISA warns to take care not to re-infect clean systems during recovery.
Handle the ransom question properly. The UK NCSC and UK law enforcement do not encourage, endorse or condone paying ransoms, and point out that payment does not guarantee you get your data back. Whatever you decide, decide it with your lawyers and insurer, check sanctions law, and never let someone negotiate informally over email. Do not contact the attacker before you have that advice.
Tell the people who need to know. Work out your obligations with your lawyers:
- Regulators and data protection authorities, if personal data may have been accessed. Some laws set tight deadlines; under UK GDPR, for example, reportable breaches must reach the ICO within 72 hours of becoming aware.
- Law enforcement and the national CERT. In Pakistan that is PKCERT; in the UK, Action Fraud and the NCSC. CISA also recommends consulting law enforcement about possible decryptors.
- Staff, with a short, honest message: what has happened, what to do and not do, and where updates will come from. People will talk anyway; give them the facts.
- Customers and partners, once you know enough to be accurate. Say what you know, what you do not know yet and when you will update them.
Plan for people, not just systems. Split the team into shifts. Tired people make the mistakes that turn containment into re-infection.
What to prepare before it happens
When we run incident response readiness work, these are the things we check first, because they are cheap to fix on a calm day and expensive on a bad one:
- A named incident lead and deputy, with authority to isolate systems without waiting for approval.
- An out-of-band contact list on paper and on personal phones: the team, the insurer's hotline, your response firm, your lawyers and your bank.
- Backups the attacker cannot reach: offline or immutable, with separate credentials, and a restore you have actually tested.
- Logs kept long enough to answer "how did they get in?", not just a few days.
- A one-page plan that covers the first hour, not a 60-page document nobody opens.
- A practised tabletop. Walk the leadership team through this scenario for an hour. Every gap you find then is one you will not find at 06:40 on a Monday.
Checklist for the first 24 hours
- Incident lead and deputy named; decision log started.
- Out-of-band communication in use; no response talk on corporate email.
- Affected systems isolated from the network, left powered on where possible.
- Backups disconnected and checked.
- Insurer notified; response firm and lawyers engaged.
- Memory and disk images taken from a sample of affected machines; logs exported.
- Scope answered: systems, accounts, entry point, data taken.
- Privileged and service account passwords reset; cloud sessions revoked;
krbtgtreset planned. - Entry point closed before any restore.
- Restore order agreed by the business; clean network segment ready.
- Notification obligations worked out with lawyers; staff briefed.
- Team on shifts.
Sources
- CISA, #StopRansomware Guide, including the ransomware response checklist.
- UK NCSC, Ransomware guidance for organisations.
- NIST, SP 800-61 Rev. 3: Incident Response Recommendations and Considerations for Cybersecurity Risk Management.
- Microsoft, AD Forest Recovery: resetting the krbtgt password.
- No More Ransom, decryption tools.
- ICO, Personal data breaches: a guide.
Written by
Arslan Amin
Chief Operating Officer, FutaqSec