Ransomware is malicious software that blocks access to computer systems or encrypts data and then demands a ransom to release it. As early as December 1989, floppy disks labelled “AIDS Information” were posted out with a Trojan horse that encrypted part of the hard disk’s root directory — Virus Bulletin (January 1990) called it an attempt to extort money from PC users. A campaign today, however, looks nothing like those early cases: encryption is the last step, not the first.
That difference decides what a sensible security audit looks like in this area. The point is not to tick off a list of tools, but to collect evidence that the controls will hold on the day the attacker already has administrative rights. This article explains how ransomware works, which controls genuinely raise the cost of an attack, and what a ransomware protection audit checklist should verify — with one decisive question at the top.
📚 Read the complete guide: Ransomware: What is Ransomware and How to Protect Yourself - Guide
Why is encryption the last step of an attack rather than the first?
Before the ransom note appears, the attacker has usually done several things: taken over an account, escalated privileges, worked out where the data lives, and dealt with the backups. CISA puts this plainly in the #StopRansomware Guide (October 2023 publication): “Ransomware actors often hunt for and collect credentials stored in the targeted environment and use those credentials to attempt to access backup solutions; they also use publicly available exploits to target unpatched backup solutions.”
The practical conclusion is uncomfortable: if encryption is the end, then every earlier stage was a moment at which the attack could have been interrupted — provided somebody was watching. The ransomware attack on Colonial Pipeline of 7 May 2021 shows the scale of the consequences: CISA recalls that images of queues of cars at petrol stations along the US East Coast went around the world at the time (CISA statement of 7 May 2023). The impact on public administration is described in the fiscal and policy note on Senate Bill 30 (SB 30, January 2020) by the Department of Legislative Services of the Maryland General Assembly: on 7 May 2019 around 7,000 Baltimore city employees lost access to their computers, the city went for weeks without its main services, and the city’s budget office put the cost of the attack at about USD 18 million (USD 10 million to restore and repair the networks and USD 8 million in potentially lost or delayed revenue). That is why the audit question is not whether the organisation has antimalware software, but how quickly somebody reacts to the signals that precede encryption — the same ones CISA describes in the guide quoted above.
Which types of ransomware are most common today?
The distinction matters for an audit, because it determines what can be recovered at all.
- Crypto ransomware — encrypts files, and recovering them without the key is effectively impossible. WannaCry is the textbook example: according to the UK National Audit Office (report of October 2017), the attack of 12 May 2017 hit over 200,000 computers across at least 100 countries.
- Locker ransomware — does not encrypt data but blocks access to the operating system. The Reveton trojan, which displays a fake law-enforcement warning, is one example — the FBI warned about it in August 2012, and Microsoft describes it as a lock screen that pretends to be from the local police.
- Ransomware-as-a-Service (RaaS) — a model in which the authors of the tooling make it available to other criminals in exchange for a share of the ransom. Cerber, which Check Point researchers described as Ransomware-as-a-Service in August 2016, is one example. RaaS did not change the mechanics of the attack; it lowered the technical barrier to entry.
On top of that comes extortion based not on encryption but on the threat of publishing stolen data. A backup does not defend against that variant — and that is the first thing an audit should say to the board without softening it.
How does ransomware get into a corporate network?
CISA organises its prevention guidance by initial access vector, and that same ordering is worth carrying into an audit, because it immediately shows whose responsibility each item is.
- Internet-facing services. The #StopRansomware Guide is categorical here: “Do not expose services, such as remote desktop protocol, on the web. If these services must be exposed, apply appropriate compensating controls to prevent common forms of abuse and exploitation.” In audit practice it is enough to compare the list of public addresses with the list of services the IT team believes are running.
- Phishing. An attachment with a macro, or a link to a site impersonating a known service, is still the cheapest way in.
- Unpatched software. Exploit kits automate the use of known vulnerabilities, so the window between a patch being published and being deployed is a real attack window.
- Suppliers and the trusted channel. Compromising a software vendor or an infrastructure maintenance provider bypasses every control placed on external traffic.
Which technical controls actually raise the cost of an attack?
Not every control carries the same weight. The ones below change the attacker’s arithmetic rather than merely filling in a box in a policy document.
- Multi-factor authentication on remote access and on administrative accounts. A stolen password stops being sufficient — and credentials are the currency of this attack.
- Least privilege and order in service accounts. An account holding domain administrator rights “because it was easier to deploy that way” is a ready-made path to the backups.
- Network segmentation. This is not about a diagram on a slide but about whether traffic between zones is genuinely blocked — and whether anyone verified it by attempting it rather than by reading the configuration.
- Vulnerability management and patching discipline, in a deliberate order: what is visible from the internet comes first.
- EDR/XDR with a real response. Endpoint telemetry is worth something only if somebody looks at it at night and at weekends too.
Does employee training still change anything?
It does, but not where it is usually measured. Training will not eliminate the click — with a good enough pretext, anyone clicks. What training genuinely delivers is a shorter interval between the click and the report.
That is why an audit should measure two things other than the share of staff who “completed the training”: whether there is a single, universally known channel for reporting a suspicious message, and how much time passes between a phishing simulation and the first report from an employee. The first number describes the procedure, the second describes the culture.
When did you last restore a production system from backup, and how long did it take?
This is the question that settles the whole audit. The rest of the checklist matters, but the answer to this one tells you whether the controls exist in reality or only on paper.
In NIST SP 1800-11, “Data Integrity: Recovering from Ransomware and Other Destructive Events” (final version, September 2020), the challenge is stated as follows: “Organizations must be able to quickly recover from a data integrity attack and trust that any recovered data is accurate, complete, and free of malware.” Both conditions carry equal weight: quickly restoring an infected state is not a restore.
How to design such a test so that its result counts as evidence:
- Pick a production system, not a test one. Restoring the marketing department’s file server says nothing about the system that invoicing depends on.
- Measure the time to service restoration, not the time to copy files. What counts is the moment a user completed the first real operation, not the moment the data landed on disk.
- Restore onto clean infrastructure, assuming the production environment is unavailable and administrative credentials are to be treated as compromised.
- Check data integrity and age. How many transactions were lost between the last good copy and the moment of encryption? That is the real business loss the board will ask about.
- Write down the record — date, elapsed time, and the name of the person who ran it. Without it, as far as an audit is concerned, the test never happened.
The first recommendation in the same CISA guide reads: “Maintain offline, encrypted backups of critical data, and regularly test the availability and integrity of backups in a disaster recovery scenario.” The word “regularly” carries operational weight here: a test performed once, at system rollout, describes a configuration from several years ago.
What separates “we have backups” from “we can restore”?
The first sentence describes a job in a scheduler. The second describes a proven organisational capability. Most of the failed restores seen after an incident live in the gap between them.
CIS states it in Controls v8.1 as Control 11, “Data Recovery”: “Establish and maintain data recovery practices sufficient to restore in-scope enterprise assets to a pre-incident and trusted state.” The last two words are the whole difference: a copy that may have been encrypted or tampered with, and nobody can say which, does not meet that criterion.
A separate question is whether the copy survives at all once the attacker holds administrative rights. The immutability mechanism and the test that settles it are broken down in the text on immutable backup for businesses; designing the restore procedure itself is the substance of backup and disaster recovery.
What should a ransomware protection audit checklist verify?
Every item below has an assigned proof — what the auditor has to see before the item counts as satisfied. The answer “yes, we have that” is not proof.
Access and identity. MFA on every remote access path — proof: a report from the identity system, not an administrator’s assurance. Privileged accounts — proof: a list of accounts with the date of the last password change and a named owner. Service accounts — proof: a register with the scope of rights and the date they were last reviewed.
Segmentation. Rules between zones — proof: the result of an attempt to pass traffic that should have been blocked. Administrative access to backup systems — proof: a demonstration that they cannot be reached from an ordinary workstation.
Backups. The 3-2-1 rule — proof: a configuration showing three copies on two media with one held offsite. Separation from the production domain — proof: a failed attempt to delete a copy using domain administrator credentials. Restore — proof: the record of the last test together with the time to service restoration.
Detection and response. Endpoint telemetry — proof: a question about a specific event from a week ago that the data can answer. Response time outside working hours — proof: the timestamp of an alert and the timestamp of the first action in a real event. Incident response plan — proof: a copy available when the network is down; the CISA guide puts it in one sentence — “Ensure a hard copy of the plan and an offline version is available.”
Which audit findings come up most often?
The four below recur regardless of industry or company size.
- Backups inside the same authentication domain as production. The domain administrator account the attacker takes over during the campaign is also the account entitled to delete the copies. The protection and the thing meant to rescue it then share a single point of failure.
- No restore test, or a test run on a non-production system. The most common answer is “the backup runs and does not report errors” — which is information about a job, not about a capability.
- Service accounts with administrative rights and no password rotation. They were usually created during some deployment years ago and nobody dared to narrow them, in case “something stops working”.
- An incident response plan nobody has rehearsed. The document exists, but it is unclear who decides to disconnect the network or who talks to customers. We lay out the bill for that omission in the article on how much ransomware costs a company.
What should you do in the first hours after detecting an attack?
Order matters here, because some natural reflexes destroy evidence or make the situation worse.
- Isolate, do not power off. Disconnecting from the network stops the spread; cutting the power wipes the contents of memory, which often holds the only trace of how the attacker got in.
- Activate the response plan and name one decision-maker. An incident in which three people make contradictory calls in parallel lasts longer than the attack itself.
- Assess the scope before restoring. Restoring data into an environment where the attacker still has access ends in a second round of encryption.
- Restore in order of service priority, into a clean segment, making sure no unverified system is added to the restored environment. That sequence follows directly from the recovery steps in the CISA guide.
- Do not pay reflexively. Payment guarantees neither the recovery of the data nor that stolen data will stay unpublished.
Who should you report the incident to in Poland, and by when?
Reporting is not a formality — it is also how you obtain support and intelligence about the campaign that hit you.
- CERT Polska. The team states on its own site (as read on 18 September 2026) that it “operates within the structures of NASK – the National Research Institute” and that “since the Act of 5 July 2018 on the National Cybersecurity System came into force, the team has performed part of the duties of CSIRT NASK, pursuant to Article 26 of that Act” (own translation from Polish). Reports are filed through the form at incydent.cert.pl. CERT Polska is not a law enforcement body — a criminal complaint is filed separately, with the prosecutor’s office or the police.
- The President of the Personal Data Protection Office (UODO). Where the incident amounts to a personal data breach, Article 33(1) GDPR requires the controller to notify the supervisory authority without undue delay and, where feasible, not later than 72 hours after having become aware of it — unless the breach is unlikely to result in a risk to the rights and freedoms of natural persons. A notification made after that deadline must be accompanied by reasons for the delay.
- The competent CSIRT, if you fall under the National Cybersecurity System Act. The scope of that duty depends on how the entity is classified, and the circle of entities covered by the act was changed by the amendment of 23 January 2026 (Journal of Laws 2026 item 252, in force from 3 April 2026). That determination is worth making before an incident rather than during one.
Where can the attack realistically be stopped?
Encryption is the last step of an attack, not the first. Before it happens, the attacker typically spends days moving through the network: taking over an account, escalating privileges, disabling endpoint protection and deleting backups. Those are precisely the actions that can be detected — provided somebody watches endpoint telemetry and responds outside working hours.
For an organisation without its own round-the-clock team that means a managed service: the scope is described under managed EDR/XDR. This text covers prevention, detection and audit; whether the copy survives the moment an attacker already holds administrative rights is a separate question — the immutability mechanism and the test that settles it are broken down in the text on immutable backup for businesses.
Related Terms
Learn key terms related to this article in our cybersecurity glossary:
- Ransomware — Ransomware is a type of malicious software (malware) that blocks access to a…
- Network Security — Network security is a set of practices, technologies, and strategies aimed at…
- Wireless Network Security — Wireless network security refers to the measures and practices used to protect…
- Cybersecurity — Cybersecurity is a collection of techniques, processes, and practices used to…
- Encryption — Encryption is the process of converting data from a human-readable format to…
Learn More
Explore related articles in our knowledge base:
- Cyber Security in the Company: Effective data protection strategies
- Cloud Migration Strategies: Best Practices, Challenges and Cost Optimization
- Cybersecurity Threats and Strategies for Local Governments - Comprehensive Guide
- Who Does the National Cybersecurity System Cover? Entities, Operators, Providers and Authorities
- What is CASB and why is it necessary for data protection in SaaS applications?
Explore Our Services
Need cybersecurity support? Check out:
- Security Audits - comprehensive security assessment
- Penetration Testing - identify vulnerabilities in your infrastructure
- SOC as a Service - 24/7 security monitoring
Related topics
See also:
- NIS2 for local government — implementation
- NIS2 Funding Calculator for Local Government
- EDR vs XDR vs NDR — comparison
Frequently asked questions
Which types of ransomware do cybercriminals use most often?
The three main types are: crypto ransomware (encrypts files and demands a ransom for the decryption key — example: WannaCry), locker ransomware (blocks access to the operating system without encrypting files) and Ransomware-as-a-Service (RaaS — a business model that makes attack tooling available to less technical criminals, example: Cerber). RaaS has lowered the technical barrier to entering this line of crime.
How does ransomware get into company systems?
The main ransomware infection vectors are: phishing attacks (malicious attachments or links in emails), remote access services exposed to the internet, including RDP, vulnerabilities in unpatched software, infected USB media and supply chain attacks. After the initial infection the attacker usually maps the network and gathers credentials before launching encryption.
What should a company do if it falls victim to a ransomware attack?
Isolate the infected systems from the network at once to stop the spread. Do not switch the computers off — memory holds evidence. Activate the Incident Response Plan (IRP), report the incident to CERT Polska through the form at incydent.cert.pl and, if a personal data breach has occurred, notify the President of the Personal Data Protection Office (UODO) under Article 33(1) GDPR. Before deciding whether to pay a ransom, contact an incident response team.
What are the most effective strategies for preventing ransomware attacks?
The most effective protection is layered: backups following the 3-2-1 rule with at least one copy outside the production authentication domain, network segmentation that limits lateral movement, vulnerability management and patching discipline, multi-factor authentication, phishing awareness training, and EDR/XDR solutions that monitor behaviour in real time. The measure of that strategy is a rehearsed restore, not the number of tools deployed.
