What ransomware became
Ransomware stopped being malware and became an industry. Understanding the business model is not background colour: it explains why the technical advice of a decade ago no longer works.
From encryption to extortion. The original model encrypted your files and sold you the key, which made backups a complete answer. The current model takes the data first and encrypts second, so that a perfect restore still leaves the attacker holding your data and a threat to publish it. The authorities state the consequence plainly: paying "will not ensure your data is decrypted, that your systems or data will no longer be compromised, or that your data will not be leaked." Three separate promises, none of them guaranteed by a payment.
The structure of the industry. The dominant arrangement separates the people who build the software from the people who use it. One group develops and maintains the ransomware and the payment infrastructure; affiliates conduct the intrusions and take a share. The joint advisory on the group that was, in its own words, "the most active global ransomware group and RaaS provider in terms of the number of victims claimed on their data leak site" in 2022 is organised around exactly that distinction, documenting the behaviour of affiliates rather than of a single actor.
Why that matters to a defender. It means the intrusion you suffer is not characteristic of a brand. Affiliates have different skill levels, different tooling and different habits, so two incidents attributed to the same name may look nothing alike. Planning around the published profile of one group is less useful than planning around the access routes in lesson two, which are common to all of them.
The supply chain of access. Intrusions are frequently bought rather than achieved. The review of the Lapsus$ attacks found that initial access brokers and the infostealer ecosystem — "whereby anyone can buy valid login credentials for a target" — were "highly effective means of initial entry". The implication is uncomfortable: your exposure includes credentials already for sale, taken from a personal device or an unrelated breach, which no control inside your network prevented.
Extortion without encryption. Some groups now skip the encryption altogether and simply steal and threaten, which removes every signal a defender was watching for. No mass file modification, no ransom note on the desktop, nothing for an anti-malware product to detect. If your detection strategy assumes encryption, it will miss this entirely, and the first notification will be the extortion itself.
What this chapter changes about the plan. Three things. Backups answer availability and no longer answer confidentiality, so data protection obligations survive a perfect restore. Detection must cover large-scale data movement outward, not only encryption. And the people who will handle the incident need to understand that the attacker's leverage is a publication threat, which is a legal and communications problem as much as a technical one.
The six ways in
The published guidance organises prevention by initial access vector, which is the right structure because it maps controls onto the way intrusions actually begin. There are six, and the list is short enough to audit yourself against.
The six, as the authorities name them. Internet-facing vulnerabilities and misconfigurations; compromised credentials; phishing; precursor malware infection; advanced forms of social engineering; and third parties and managed service providers.
Internet-facing vulnerabilities. Not a theoretical category. The LockBit advisory lists specific exploited vulnerabilities by identifier, including a managed file transfer product, a print management product, and the logging library whose flaw became general knowledge in 2021. The pattern is consistent: edge devices and file transfer appliances, because they are exposed by design, often unpatched because they are treated as appliances rather than servers, and valuable because they handle data in bulk.
Compromised credentials. The same advisory records that affiliates "may leverage VPN or RDP brute force credentials as an initial access" route and that they "exploit RDP to gain access to victims' networks". Remote access exposed to the internet without phishing-resistant authentication is the single most reliable way in, and it is usually present because somebody needed it urgently in 2020 and nobody removed it.
Phishing and precursor malware. Phishing remains a primary route, and precursor malware is the category defenders most often dismiss. A commodity infostealer or loader on one machine is not a minor alert to be cleaned and closed; it is frequently the first stage of an intrusion that will be sold on. Treating a precursor infection as an incident rather than as hygiene is one of the higher-value changes available to a security operation.
Advanced social engineering. Version 3.0 of the guidance, published in October 2023, specifically added material about threat actors impersonating employees. That is the same technique the identity course documents in detail — persuading a service desk to reset a password or move an authenticator. It belongs in a ransomware course because it is now a ransomware access route, not a separate discipline.
Third parties and managed service providers. The route with the widest blast radius, because a provider with administrative access to many clients is a single compromise with many victims. If a supplier holds remote access into your estate, their authentication is your authentication, and you are entitled to ask what it is.
How to use this list. Take each of the six and write down, in one sentence, the control you rely on and the evidence that it is working. The exercise usually produces two or three sentences that cannot be completed, and those are your actual exposure. It is a more honest assessment than a maturity score, and it takes an afternoon.
Before the encryption
Encryption is the end of the intrusion, not the beginning. By the time files start changing, the attacker has usually been present for some time and has done specific things that are both detectable and, if you detect them, decisive.
What the interval is for. An attacker who encrypts immediately gets one machine. An attacker who waits gets the whole estate, and the waiting is spent on a recognisable sequence: establishing persistence, finding and stealing credentials, mapping where the valuable data is, moving it out, and locating and destroying the ability to recover. The last of those is the step that converts an incident into a crisis.
Credential theft inside the network. The guidance notes that ransomware actors "often hunt for and collect credentials stored in the targeted environment". This is why lesson two's compromised-credential vector is not only an entry problem: the credentials harvested after entry are what turn one foothold into estate-wide access. Everything the identity course says about standing privilege applies with particular force here, because a stolen administrator credential is what makes mass encryption possible at all.
The deliberate attack on recovery. Attackers look for backups and remove them, because an organisation that can restore does not pay. The guidance states the motivation directly — actors "attempt to find and subsequently delete or encrypt accessible backups to make restoration impossible unless the ransom is paid." The operative word is accessible: a backup reachable using credentials the attacker now holds is not a backup, it is another copy of the target. Lesson six is what to do about that.
Data movement outward. Before the encryption comes the exfiltration, and it is the most reliable detection opportunity in the whole sequence because it is hard to disguise and easy to measure. Version 3.0 of the guidance added a specific indicator for exactly this: abnormal amounts of data outgoing over any port. Not over the ports you monitor — any port. Most organisations instrument inbound traffic carefully and measure outbound volume not at all, which means the loudest signal in the attack passes unexamined.
Why this lesson is the most actionable one. Every step in the interval is an opportunity, and each is easier to detect than the intrusion itself. A new administrative account, a credential-dumping tool, an unusual authentication from a service account, a backup job that failed or was deleted, a sudden outbound transfer of tens of gigabytes — any one of these is a chance to intervene before the ransom note. The organisations that contain these incidents cheaply are, almost without exception, the ones that caught something in this window.
The honest caveat. The interval is not guaranteed and it is shrinking. Some intrusions move from access to encryption in hours, and an attacker who bought working credentials may skip several steps entirely. Design detection for the interval, but do not let the plan depend on its length.
Reading the numbers honestly
Ransomware is the subject where published statistics are least reliable and most confidently quoted. This lesson is about reading them properly, because a decision justified with a wrong number is a wrong decision that nobody can audit.
A worked example from a primary source. The FBI's Internet Crime Complaint Center reported, for 2024, total losses across all crime types of $16.6bn, and recorded 3,156 ransomware complaints with adjusted losses of $12,473,156. Those two figures sit in the same document, and the second one is obviously too small — about 0.08% of the total, for the crime type the same report calls the most pervasive threat to critical infrastructure.
The report explains why, which is the valuable part. Its own footnote states that the ransomware loss figure "does not include estimates of lost business, time, wages, files, or equipment, or any third-party remediation services acquired by an entity." It adds that "in some cases, entities do not report any loss amount to FBI, thereby creating an artificially low overall ransomware loss rate", and that the number "only represents what entities report to FBI via IC3". So the figure is not wrong; it measures something much narrower than the phrase "ransomware losses" suggests.
A second example, from a different authority. The joint letter from the UK's data protection regulator and its technical cyber authority cites a cost estimate of £1.1bn from computer misuse incidents against individuals in England and Wales, and immediately qualifies it: "this is a partial estimate only. Crucially, this does not include the cost to businesses which are thought to bear the majority of the cyber crime costs, meaning the true cost from cyber crime will be much higher."
The pattern worth internalising. Both authorities published a number and then told you what it omits. The failure is not theirs; it belongs to everyone who quotes the figure without the qualification. When you see a ransomware statistic in a vendor deck or a board paper, the questions are always the same: who reported it, what counted as a loss, who was in the sample, and what the source says about its own limits.
What the figures can legitimately support. Direction and relative scale. The IC3 report states ransomware complaints rose 9% from 2023 and that ransomware was "again the most pervasive threat to critical infrastructure". A European threat landscape report published in October 2025 analysed 4,875 incidents across the twelve months from July 2024. Counts of incidents and reported trends are defensible. Totals of financial loss are not, and should not appear in a business case without the qualification attached.
The number you can actually produce. Your own. What would a week of downtime in your most critical process cost, in revenue, in regulatory exposure and in staff time? That figure is estimable from facts you hold, it is specific to the decision in front of you, and it is far more persuasive than an industry total that the publishing authority has itself disowned.
The controls that actually bear on it
There is no ransomware control. There is a set of ordinary security controls whose presence or absence decides how a ransomware intrusion goes, and this lesson is which ones, in the order their absence hurts.
Why a dedicated ransomware product is the wrong first question. Ransomware is an outcome, not a technique. The intrusion that produces it uses the access routes in lesson two and the escalation in lesson three, all of which are addressed by controls you already have names for. An organisation with unpatched edge devices, remote access without phishing-resistant authentication and no outbound monitoring will not be saved by a product with ransomware in its name.
The controls, ranked by how often their absence is decisive.
- Phishing-resistant authentication on anything reachable from the internet, and especially on remote access. The documented intrusions in this course begin with credentials more often than with exploits, and credentials can be bought.
- Patching the exposed edge, on a timetable that treats appliances as servers. The vulnerabilities named in the LockBit advisory were in file transfer and print management products — the category most likely to be owned by nobody in particular.
- No standing administrative privilege. Mass encryption requires estate-wide rights. An attacker who steals an ordinary account and finds no path to permanent administrative credentials has a much smaller incident to offer you.
- Outbound volume monitoring, because exfiltration is the loudest moment in the attack and the one most organisations do not instrument.
- Treating precursor malware as an incident, not as a cleaned alert.
- Recoverability that an attacker with your credentials cannot reach, which is the next lesson.
The published baselines, and their honest limits. Freely available performance goals state specific, checkable actions — for example a minimum password length of 16 or more characters for password-protected assets where technically feasible, and secure offsite and offline storage of backups with recovery tested "no less than once per year". Configuration benchmarks give versioned baselines per platform. These are useful because they are external and specific; they are not a strategy, and an organisation can satisfy a checklist while leaving one of the six access routes wide open.
On frameworks. The relevant profile for this subject was reissued in June 2026 as a Cybersecurity Framework 2.0 community profile for ransomware risk management, replacing a version built on the older framework. A profile of this kind is valuable for structuring a programme and for talking to a board. It will not tell you which of your own systems is most exposed today, which is what lesson two's one-sentence-per-vector exercise is for.
The hardening nobody does. The October 2023 revision of the guidance added hardening advice for a specific file-sharing protocol and for web browsers. Both are unglamorous, both are standard in most estates, and both are routinely skipped because nothing visibly breaks when they are. The work that prevents these incidents is almost entirely of this character, which is why it needs to be scheduled rather than motivated.
Recoverability an attacker cannot reach
Backups are covered in every security course. This lesson is about the specific property that matters against an adversary rather than against a disk failure, and about the two stages organisations discover too late.
The adversarial requirement. A disaster recovery design assumes failure, which is indifferent to your credentials. Ransomware assumes malice, which uses them. The guidance makes the distinction operational: backups must be maintained offline, because actors "attempt to find and subsequently delete or encrypt accessible backups". Anything an attacker holding your administrative credentials can delete is not a recovery capability.
The three requirements in the guidance's own sentence. "Maintain offline, encrypted backups of critical data, and regularly test the availability and integrity of backups in a disaster recovery scenario." Offline, encrypted, and tested. The third is the one that fails, and it fails silently: a backup that has never been restored is a hypothesis about a file format, a retention setting and a credential, all of which can be wrong at once.
The step that is specific to malware and almost always missing. Scan backups for malware before restoring. The UK guidance states it directly, and the reason is that the attacker was present before the encryption — so a backup taken during the dwell period may contain their persistence. An organisation that restores from the most recent clean-looking backup can restore the intrusion along with the data, which is how second encryptions happen a week after recovery.
What else needs backing up. Not only the data. The guidance says to keep offline copies of infrastructure-as-code template files so resources can be redeployed quickly, and to store applicable source code or executables with offline backups, noting that rebuilding from system images is more efficient but "some images will not install on different hardware or platforms correctly". Add the identity provider's configuration, the detection rules, and the key material without which restored data is ciphertext.
On immutability, with the source's own caution. Provider immutable storage can protect data without a separate environment, and the guidance pairs that with a warning rarely repeated by sellers: it "does not meet compliance criteria for certain regulations and misconfiguration can impose significant cost." Data you cannot delete is data you cannot delete when a data subject or a regulator requires it, and a retention lock set wrongly is an unstoppable bill. Set the period deliberately, with someone who has read both the threat model and the retention obligations.
The cloud trap, stated once because the cloud course covers it fully. Synchronisation is not backup: encrypted local files sync upward and can overwrite good copies. If your recovery plan depends on a file-sync service, test it against encryption specifically rather than against deletion.
The test that settles it. Assume an administrator identity has been compromised for three weeks. From there, what can you restore, from where, with which credentials, in how long, and how do you know the restored image is clean? Most organisations find their recovery plan and their attack surface share a credential, and that the answer to the cleanliness question is that nobody has one.
The first hour
The value of an incident response plan is almost entirely in the first hour, and the published checklist is specific about sequence rather than merely about content.
The instruction that is easy to miss. The response checklist carries an explicit note to "move through the first three steps in sequence." Incident response guidance is usually a list; this is an order. The reason is that the first steps are about preserving the ability to decide — isolate, then establish scope, then record — and doing them out of order destroys information you will need.
Determine what is affected, and isolate it. The checklist's first step, with a hierarchy that is more practical than most plans contain. If several systems or subnets appear affected, "take the network offline at the switch level", because disconnecting individual machines during an incident may not be feasible. Prioritise isolating the critical systems essential to daily operations. If taking the network down is not immediately possible, physically unplug affected devices or remove them from the wireless network. And for cloud resources, "take a snapshot of volumes to get a point in time copy" — evidence preservation at the moment it is cheapest.
Why the switch-level instruction matters organisationally. Somebody has to be authorised to take the network down at two in the morning without seeking permission. If that authority is not delegated in advance, the delay while it is sought is the window in which the encryption completes. This is a governance decision disguised as a technical one, and it should be settled in writing before it is needed.
What the first hour is not for. Rebuilding, negotiating, or determining attribution. It is for containment, evidence and notification. The decisions in lessons eight to eleven — whether to pay, whom to tell, what the sanctions position is — all go better with evidence and worse under time pressure, and the first hour is what buys the time.
Notification, which has deadlines you do not control. Regulatory reporting obligations run on clocks set by law rather than by your investigation. The UK letter in lesson ten notes there "may be a regulatory requirement to report to the ICO as the data regulator", and separately that reporting to the technical authority and to law enforcement serves a different purpose. Know which of your obligations are triggered by suspicion rather than by confirmation, because those are the ones that fall inside the first day.
The modern framing of response. The current national guidance on incident response, revised in 2025, is written as a community profile of a risk management framework rather than as a standalone lifecycle — that is, response treated as part of continuous risk management. For ransomware specifically that framing is apt, because almost everything that determines how the first hour goes was decided months earlier: the logging tier, the backup design, the delegated authority, and whether anyone had rehearsed.
The rehearsal worth more than the document. Walk one scenario end to end with the actual people. Who declares the incident? Who may disconnect the network? Who contacts law enforcement, and do they have the number? Who decides about a payment, and do they know about lesson nine? The gaps that surface in that hour are the ones that would otherwise surface during the incident.
The decision about paying
This lesson sets out what is actually known, because this is the decision on which people most want certainty and have least, and because the organisations that handle it worst are the ones encountering the question for the first time at speed.
What the authorities say, and in what register. The position is consistent across jurisdictions and it is advisory rather than prohibitive. The four authoring agencies of the main guidance "do not recommend paying ransom". The UK's technical authority states that "law enforcement do not encourage, endorse, nor condone the payment of ransom demands." The US government "strongly discourages all private companies and citizens from paying ransom or extortion demands". None of these is a prohibition, and understanding that distinction is the start of handling the question honestly.
What a payment does not buy. Three separate things, each stated in the guidance: that your data will be decrypted; that your systems and data will no longer be compromised; and that your data will not be leaked. The UK guidance adds two more considerations — "there is no guarantee that you will get access to your data or computer" and "you're more likely to be targeted in the future."
The reasoning that is specific to the current model. Because the data was taken before the encryption, a payment buys a promise of deletion from a criminal enterprise, verifiable by nothing. If your exposure is a publication threat rather than an availability problem, the payment does not restore a state of affairs; it purchases an assurance. That is a materially different transaction from buying a decryption key, and it should be described that way to whoever is deciding.
What does sometimes work. Law enforcement holds decryption keys. The FBI's 2024 report records that since 2022 it has "offered up thousands of decryption keys to victims of ransomware, avoiding over $800 million in payments." That is an argument for contacting law enforcement early, before a payment decision, and it is one of the more concrete reasons to make that call rather than to treat it as a formality.
The honest statement of the other side. There are circumstances in which organisations pay: lives or safety depend on a system, recovery is impossible because the backup design failed, or the publication of the data would be more damaging than any alternative. Pretending otherwise produces guidance nobody follows. What the authorities' position means in practice is that a payment should be a deliberate, documented, advised decision with the limits above understood — not a default reached at three in the morning by whoever is available.
Who should decide, and what they need in front of them. Not the security team alone. The decision needs the executive who owns the consequences, legal advice, and — crucially — the sanctions position in the next lesson, because that is the part most likely to be missed and the only part that can convert a commercial judgement into a legal violation. Settle in advance who holds this decision and what they must be shown before making it.
Why a payment can be a sanctions matter
This is the lesson most ransomware training omits, and it is the one that can turn a defensible commercial decision into a legal violation. The guidance itself makes the link: the response checklist notes that "paying ransoms may pose sanctions risks" and refers the reader directly to the sanctions authority's advisory.
What the advisory is. An Updated Advisory on Potential Sanctions Risks for Facilitating Ransomware Payments, issued by the US Treasury's Office of Foreign Assets Control on 21 September 2021. It is explicitly "explanatory only and does not have the force of law" — but it describes how the authority will approach enforcement under laws that do.
Who is exposed, which is wider than the victim. The advisory states that companies facilitating payments on behalf of victims — "including financial institutions, cyber insurance firms, and companies involved in digital forensics and incident response" — not only encourage future demands but "may risk violating OFAC regulations". So the insurer, the incident response firm and the bank processing the transfer are all in scope, not merely the organisation that was attacked.
The sentence that changes how this decision must be taken. "OFAC may impose civil penalties for sanctions violations based on strict liability, meaning that a person subject to U.S. jurisdiction may be held civilly liable even if such person did not know or have reason to know that it was engaging in a transaction that was prohibited under sanctions laws and regulations administered by OFAC."
Read that carefully, because it removes the defence people assume they have. Not knowing who received the money is not an answer. The attacker will not tell you, attribution is uncertain, and the standard is strict liability — so the risk cannot be managed by ignorance, only by due diligence and by the mitigating conduct in the next lesson.
Why this is hard in practice. You are making a payment to an unidentified recipient, under time pressure, in a form that resists tracing, to a party who may or may not be designated or connected to a designated person. There is no process that makes that comfortable. What exists is a route: check before paying, involve counsel, contact the relevant authorities, and document the enquiry. The advisory provides contact routes precisely because the expectation is that you will use them if there is "any reason to suspect the cyber actor demanding ransomware payment may be sanctioned or otherwise have a sanctions nexus."
The range of outcomes. Enforcement responses run from non-public ones — a No Action Letter or a Cautionary Letter — to public civil monetary penalties. Which end you land at is substantially influenced by conduct, which is the subject of the next lesson and is the most practically useful finding in this course.
For readers outside US jurisdiction. The same structural issue arises under other sanctions regimes. The UK joint letter states that "while payments are not usually unlawful, payers should be mindful of how relevant sanctions regimes (particularly those related to Russia) — and their associated public guidance — may change that position." The specific designations differ; the principle that a payment may be a separate legal matter from the attack does not.
What regulators say they will credit
Two authorities in two jurisdictions have put in writing what they will and will not treat as mitigation after a ransomware incident. Read together they produce the most useful finding in this course, and it is not the one people expect.
The sanctions authority's list of mitigating factors. Meaningful steps taken to reduce the risk of extortion are "considered a significant mitigating factor in any OFAC enforcement response." The steps it names: "maintaining offline backups of data, developing incident response plans, instituting cybersecurity training, regularly updating antivirus and anti-malware software, and employing authentication protocols, among others."
And on conduct after the attack. Reporting to the relevant authorities "as soon as possible after discovery of an attack" is treated as "a voluntary self-disclosure and a significant mitigating factor". Full and ongoing cooperation with law enforcement — "providing all relevant information such as technical details, ransom payment demand, and ransom payment instructions as soon as possible" — is also a significant mitigating factor. The advisory then says it plainly: OFAC "would be more likely to resolve apparent violations involving ransomware attacks with a non-public response" when the affected party took those steps, "particularly reporting the ransomware attack to law enforcement as soon as possible and providing ongoing cooperation."
The data protection regulator's position, in its own words. A joint letter dated 7 July 2022, signed by the UK Information Commissioner, John Edwards, and the Chief Executive of the National Cyber Security Centre, Lindy Cameron, was written because "a belief persists that payment of a ransom may protect the stolen data and/or result in a lower penalty by the ICO should it undertake an investigation. We would like to be clear that this is not the case."
It continues: "For the avoidance of doubt the ICO does not consider the payment of monies to criminals who have attacked a system as mitigating the risk to individuals and this will not reduce any penalties incurred through ICO enforcement action."
And what that regulator will credit instead. "Where the ICO will recognise mitigation of risk is where organisations have taken steps to fully understand what has happened and learn from it, and, where appropriate, they have raised their incident with the NCSC, reported to Law Enforcement via Action Fraud, and can evidence that they have taken advice from or can demonstrate compliance with appropriate NCSC guidance and support."
The finding, stated once. Both authorities, independently, credit the same two things — preparation before the attack and cooperation after it — and neither credits the payment. That has a direct consequence for how you justify security spending: offline backups, an incident response plan, training and strong authentication are not only technical controls. One authority names them as mitigating factors in an enforcement action, and the other requires evidence of compliance with national guidance. The untested backup is now a legal exposure as well as an operational one.
One practical reassurance from the letter. "Neither the NCSC nor Law Enforcement share information on incidents with any regulators without permission from the affected organisation." The fear that calling the technical authority will trigger regulatory action is addressed directly, and it removes a common reason organisations delay the call that both authorities say they will credit.
How the money moves
Ransomware payments pass through the financial system, which makes them visible to institutions that know what to look for. This lesson is for anyone in a bank, payments firm or crypto business, and it is also useful to a victim, because it explains why your payment attempt may be stopped.
The relevant advisory. FinCEN's advisory FIN-2021-A004, dated 8 November 2021, on ransomware and the use of the financial system to facilitate ransom payments. It exists because payments move through regulated institutions and because those institutions have reporting obligations that are triggered by suspicion rather than by proof.
The framing it insists on. "As no single financial red flag indicator is indicative of illicit or suspicious activity, financial institutions should consider the relevant facts and circumstances of each transaction, in keeping with their risk-based approach to compliance." This is the same discipline as any other financial crime typology: indicators are inputs to judgement, not triggers for automatic action.
Indicators the advisory names. Among them:
- A financial institution or its customer detects IT enterprise activity connected to ransomware cyber indicators or known cyber threat actors, which "may be evident in system log files, network traffic, or file information."
- A customer provides information, when opening an account or in other interactions, "that a payment is in response to a ransomware incident."
- A customer's convertible virtual currency address, or an address they transact with, "is connected to ransomware variants, payments, or related activity", whether that connection appears in open sources or in commercial or government analyses.
- An irregular transaction between an organisation — "especially an organization from a sector at high risk for targeting by ransomware (e.g., government, financial, educational, healthcare)" — and a digital forensics or incident response firm, "especially one known to facilitate ransomware payments."
What that last one means for both sides. For a bank, a sudden large payment from a hospital to an incident response firm is a legitimate thing to look at. For a victim, it means your payment route is itself an indicator, and your intermediary's relationships are part of your risk. If you are retaining a firm that will handle a payment, their regulatory posture is a question worth asking before you need them.
Why a cyber course covers this. Because the two disciplines meet here and usually do not speak. The security team sees an availability crisis; the financial crime team sees an unexplained payment to a high-risk counterparty; the sanctions question from lesson nine sits between them. An organisation whose incident plan does not involve its financial crime function will discover the connection at the worst moment, when a payment is declined or an obligation is missed.
The practical step. Make sure the people who would file a suspicious activity report (SAR; the FATF standards and many other countries call the same filing a suspicious transaction report, or STR) and the people who would handle the intrusion know each other's names before the incident. It costs one meeting and it resolves a problem that otherwise surfaces under maximum pressure.
Recovery, and what to rehearse
Recovery is where the cost is actually incurred, long after the news coverage, and it is the phase organisations plan least. This lesson is what the work consists of and what to rehearse now.
Why recovery is longer than the technical restore. Restoring data is one task among many. The others: establishing that restored systems are clean, which lesson six's scan-before-restore step addresses; rebuilding identity, because credentials harvested during the intrusion must be assumed compromised and that includes service accounts and keys; re-establishing trust with customers and counterparties; meeting regulatory and contractual notification obligations; and working through what was taken rather than what was encrypted, because the data protection consequences follow the exfiltration, not the encryption.
The order problem. Everything cannot be restored at once, and the sequence matters in ways that are not obvious under pressure. Identity usually has to come first, because nothing else can be administered without it. Then the systems that other systems depend on. Then the business processes in order of consequence rather than of loudness. A restore order decided during the incident will be decided by whoever is shouting, and it will be wrong.
The identity rebuild, which is routinely underestimated. If an attacker held administrative credentials, then every credential they could reach is suspect: user passwords, service account credentials, API keys, certificates and any signing material. The key management guidance exists precisely for the question of what to do when a key's protection may have failed, and an organisation that has never rehearsed a rotation will be attempting its first one during its worst week. Lesson four of the cloud course makes the same point about cryptoperiods; here it becomes operational.
What the published goals ask for, as a minimum. Secure offsite and offline backup storage, with backups and recovery tested on a recurring basis and "no less than once per year". Once a year is a floor, not a target, and testing a restore is not the same as testing a recovery — the first proves a file comes back, the second proves a business process runs.
The rehearsal that produces the most value per hour. Not a tabletop discussion of a scenario. A timed, partial, technical exercise: pick one business-critical service, assume its data and its administrative credentials are both compromised, and actually restore it into an isolated environment from the offline copy. Measure how long it took, what was missing, who had to be woken, and whether the restored system was verifiably clean. Every organisation that does this learns something expensive and learns it cheaply.
The four things to settle before an incident, drawn from the whole course. Who is authorised to disconnect the network without seeking permission. Who decides about a payment, and the requirement that they be shown the sanctions position first. Which of your reporting obligations are triggered by suspicion rather than confirmation. And whether your most recent restore test was a file or a business process.
A closing observation. Almost nothing in this course required a capability that does not exist or a budget that is unavailable. The controls that decide these incidents are patching an exposed appliance, authentication that resists phishing, no permanent administrative rights, measuring outbound data, an offline backup that has been restored at least once, and a delegated authority to pull the cable at two in the morning. Two separate regulators have now said in writing that they will credit exactly that kind of preparation — and that they will not credit the payment.
Fifty questions
Answer all fifty questions, then check your answers. You need 45 out of 50 to pass and receive a certificate. If you score less, you will see which answers were right and wrong, and then go through the course again before you retake the check. Your answers, progress and times are kept only in this browser.
Your answers
Your certificate of completion
Enter your name as you want it to appear, then save the certificate as a PDF. In the print window, choose Save as PDF. A certificate is issued once per completion of the course.
Rate this course
How would you rate this course? Your rating is anonymous — just a number, with no name and no account.
Bring this to your team
Your colleagues can take this course too — it is free, needs no sign-up, and nothing they enter leaves their browser. If your institution wants the same ground covered as structured training for a team, with an instructor and your own cases, we run courses in a classroom or online for groups of ten or more.
Talk to us about training your team → See the paid courses →
Saolix does not record who takes this course, so it cannot verify these certificates. The certificate confirms completion of a free self-paced course and is not an accredited qualification.
Standards and frameworks covered
The published frameworks this course teaches toward. Saolix is not affiliated with, endorsed by or accredited by any of these bodies; naming a standard is a citation, not a claim about its author.
- NIST — NIST IR 8374 Revision 1: Ransomware Risk Management, a Cybersecurity Framework 2.0 Community Profile, NIST SP 800-61 Revision 3: Incident Response Recommendations and Considerations for Cybersecurity Risk Management, NIST SP 800-57 Part 1 Revision 5: Recommendation for Key Management
- ENISA — ENISA Threat Landscape 2025
- MITRE ATT&CK — ATT&CK cloud matrix
Sources
The official documents this course relies on. Laws and guidance change, so check the current version.
- #StopRansomware Guide, version 3.0 · CISA, Multi-State Information Sharing and Analysis Center, National Security Agency and FBI
- Updated Advisory on Potential Sanctions Risks for Facilitating Ransomware Payments · Office of Foreign Assets Control, US Department of the Treasury
- FIN-2021-A004: Advisory on Ransomware and the Use of the Financial System to Facilitate Ransom Payments · Financial Crimes Enforcement Network, US Department of the Treasury
- Joint ICO and NCSC letter to The Law Society and The Bar Council · Information Commissioner's Office and UK National Cyber Security Centre
- Solicitors urged to help stem the rising tide of ransomware payments · UK National Cyber Security Centre
- Mitigating malware and ransomware attacks · UK National Cyber Security Centre
- Understanding Ransomware Threat Actors: LockBit (advisory AA23-165A) · CISA, FBI and international partners
- 2024 Internet Crime Report · Internet Crime Complaint Center, US Federal Bureau of Investigation
- NIST IR 8374 Revision 1: Ransomware Risk Management, a Cybersecurity Framework 2.0 Community Profile · US National Institute of Standards and Technology
- Cybersecurity Performance Goals 2.0 · US Cybersecurity and Infrastructure Security Agency
- NIST SP 800-61 Revision 3: Incident Response Recommendations and Considerations for Cybersecurity Risk Management · US National Institute of Standards and Technology
- NIST SP 800-57 Part 1 Revision 5: Recommendation for Key Management · US National Institute of Standards and Technology
- Review of the Attacks Associated with Lapsus$ and Related Threat Groups · Cyber Safety Review Board, US Department of Homeland Security
- ENISA Threat Landscape 2025 · European Union Agency for Cybersecurity
- CIS Benchmarks · Center for Internet Security
- ATT&CK cloud matrix · The MITRE Corporation
- Cloud security guidance · UK National Cyber Security Centre
- SS2/21: Outsourcing and third party risk management · Bank of England, Prudential Regulation Authority



