

The previous publication, “The Attack Did Not Wait for the Meeting to Finish”, followed a containment recommendation through a meeting while the attack kept moving. Here, your team has already made a decision and now has evidence that changes the grounds for it.
A response plan gives your incident team a starting point. When evidence overturns an assumption behind a live containment decision, record the fact that changed, the amended restriction, who authorised it and what the next check must establish. The British Library’s own account shows how a response can start as planned while the service and recovery problem grows beyond the original scope.
Our next publication, the special edition “Malvertising: The Website Was Trusted. The Code Was Not.”, examines how an ordinary visit to a trusted website exposed readers to malicious code delivered by a shared advertising tag.
At 08:12, your incident lead opens the plan and restricts one administrator session. The evidence points to a single account, and the rest of the service can keep running.
At 08:19, the identity team finds that a second live session used the same credential to reach another service. The restriction is still in place. The assumption that made it sufficient has expired.
A plan has to simplify. Without structure, people would begin every incident with a blank page. It therefore sets roles, escalation paths, communications, technical actions, decision thresholds and recovery priorities. That is necessary work.
Plans usually assume that the assets listed, the people available and the dependencies last tested still match the organisation on the day. New suppliers, identity connections and changes to shared services can make those assumptions out of date before anyone opens the plan.
The useful detail is the assumption behind the action you have already taken. If your incident lead restricts one session because all known activity points to it, put that premise beside the decision. The person reading the incident log can then recognise evidence that calls it into question.
The British Library’s review describes rapid crisis-plan activation and a longer reassessment of services and dependencies.
NIST SP 800-61 Rev. 3 integrates incident response with cybersecurity risk management and treats lessons as inputs to improvement across the response. Together they support making the reasons for consequential decisions visible as evidence changes.
Sources and further reading:
British Library - LEARNING LESSONS FROM THE CYBER-ATTACK: British Library cyber incident review
https://cdn.sanity.io/files/v5dwkion/production/99206a2d1e9f07b35712b78f7d75fbb09560c08d.pdf
NIST - Incident Response Recommendations and Considerations for Cybersecurity Risk Management: A CSF 2.0 Community Profile
https://csrc.nist.gov/pubs/sp/800/61/r3/final
Who and when
On 28 October 2023, ransomware hit the British Library, the UK’s national library. Researchers rely on its catalogue and ordering systems to find and request material from its collections. A searchable version of the main catalogue returned on 15 January 2024.
What happened?
The Library’s review identified a remote-access server running Terminal Services as the likely point of entry. The server lacked multifactor authentication. The Library found signs of activity from 25 October. The attackers moved through its network, copied about 600 GB of files, encrypted systems and destroyed servers, making recovery harder. The Library activated its crisis plan at 09:15 on 28 October. The National Audit Office says the Rhysida group demanded 20 bitcoin, worth about £600,000 at the time, and published stolen data on the dark web after the Library declined to pay.
Operational impact
The attack took the website, main catalogue and electronic ordering offline and cut access to online journals, databases and digital theses. The Reading Rooms stayed open, and cloud-based payroll and finance continued to work. Staff handled some requests manually. The January catalogue let readers search again, while only about half the physical collection was accessible by volume as of March 2024. In December 2025, the new main catalogue let readers sign in and place requests directly again.
Confirmed consequences
The stolen files included records from the Library’s personnel team and personal information about staff and registered readers. In March 2025, the Cabinet Office told Parliament that the response had cost an estimated £6 million to £7 million, equivalent to more than 40% of the Library’s financial reserves.
Why this case matters
The crisis plan was in motion that morning. Once the Library saw the damage to its older systems, it had to decide which services staff could run manually and which would have to wait for new infrastructure. That decision determined what researchers could still consult while the Library rebuilt.
Sources and further reading:
British Library - LEARNING LESSONS FROM THE CYBER-ATTACK: British Library cyber incident review
https://cdn.sanity.io/files/v5dwkion/production/99206a2d1e9f07b35712b78f7d75fbb09560c08d.pdf
National Audit Office - Government cyber resilience
https://www.nao.org.uk/wp-content/uploads/2025/01/government-cyber-resilience.pdf
House of Commons - Oral evidence: Government Cyber Resilience, HC 643
https://committees.parliament.uk/oralevidence/15484/html/
British Library - Cyber-attack recovery update
https://www.bl.uk/about/cyber-attack
British Library - Introducing our new catalogues
https://www.bl.uk/stories/news/introducing-our-new-catalogues
British Library - Reading Rooms and study spaces
https://www.bl.uk/visit/reading-rooms
Return to 08:19. The incident lead revises the restriction to cover the second session and records why it now falls within scope. The service owner checks whether routine work can continue through the known fallback. Your identity team verifies that both sessions have lost access and checks whether the same credential is still in use elsewhere. The record names the lead who approved the change and the person who will bring the verification result back.
Write the next review against an observable result. If both sessions are closed and the identity team finds further use of the same credential, the original decision needs another look. The record gives whoever is on duty enough context to make that call.
Run an exercise through one containment decision, then introduce a fact that invalidates its premise. Hand the incident record to someone who missed the first discussion. Ask them what changed, which action is now authorised and which piece of evidence would cause them to reopen it again.
If they need the people who made the decision to explain it, look at what was left out of the record. The gap could be the reason for the first restriction, the evidence behind the revision or the owner of the next check.
The next shift should be able to tell why your team chose a restriction and which new fact would require it to be reopened.
Our platform gives your team ways to interrupt supported malicious activity while it verifies what has changed. Ransomware Containment can automatically isolate the user or session responsible for active encryption detected on protected SMB and NFS file services. Server Intrusion Protection can interrupt supported malicious activity involving compromised administrator access, and Virtual Server Protection extends protection to supported VMware and Hyper-V environments.
Your incident lead can use the observed effects of a containment action when reassessing the wider response.
Our next special, “Malvertising: The Website Was Trusted. The Code Was Not.”, takes that question into the browser. If an approved advertising tag starts serving malicious code, how will your team establish which visits were at risk and switch off the tag while keeping the site available?
Wednesday 18 Nov 2026, Leadership-series #11: "The Organisation Itself Becomes The Bottleneck"
Friday 20 Nov 2026, Operational Attack-series #11: "The Extortion Trap Begins Before the Ransom Note"
Monday 23 Nov 2026, Resilience-series #12: "When Ransomware Becomes a Life-Critical Decision"