What ransomware is and how it gets in

Ransomware is malware that encrypts the organization’s data and demands payment in exchange for decrypting it. In most cases today, attackers also copy data before encrypting it and threaten to publish it (double extortion). So even a business that managed to restore from backup still faces exposure of customer information.

Small businesses are hit mainly because they are convenient targets: fewer layers of protection, less monitoring, and often nobody looking at the alerts. Most attacks start through one of a few familiar doors:

  • Impersonating email (phishing): a link or file that gets an employee to enter a password or run code.
  • Leaked passwords (credential stuffing): a password that appeared in a leak from another website and is also used for email or remote access.
  • Exposed remote access: remote desktop (RDP) or a management interface open directly to the internet.
  • Systems that were not updated: known vulnerabilities in servers, in the firewall or in the VPN component.

The steps below address exactly these doors. None of them guarantees there will be no attack. Together they greatly reduce the chance it will succeed, and the damage if it does.

What happens the day after

The direct damage is familiar: computers will not open, files are encrypted and work stops. What is less expected is how many things have to be done in parallel. You need to understand what was hit and where the attacker got in, stop the spread (containment), decide whether and from where to restore, and talk to customers who are waiting for service and asking whether their information leaked.

When personal information is involved, there is also a regulatory side. The data security regulations under the Privacy Protection Law require, in certain cases, reporting a severe security incident to the Privacy Protection Authority. Amendment 13 to the law, which came into force in August 2025, gave the Authority broader enforcement powers, including administrative fines. A business that cannot show it took reasonable security measures is in a harder position with the Authority and with its customers.

Paying the ransom does not solve all of this. There is no certainty the key will work, and there is no way to verify that the stolen data was really deleted.

Permissions and accounts

Ransomware reaches only what the breached user can touch. That makes limited permissions (least privilege) one of the most effective steps, and most of them do not require buying any product:

  • Day-to-day work in a regular account, and a separate administrator account only for management tasks.
  • A different local administrator password on every computer, for example with LAPS, so a breach of one computer does not open all of them (lateral movement).
  • Access to shared folders by role (RBAC), instead of “everyone sees everything.”
  • Multi-factor authentication (MFA) for email, remote access and every administrator account, including the accounts of external suppliers.
  • Immediately closing the accounts of employees who have left (offboarding).

In organizations with more servers, it is also worth considering privileged access management (PAM), in which administrator rights are granted for a limited time (just-in-time) and documented.

Updates and internet exposure

A large share of attacks exploit vulnerabilities for which a fix has already been published. Regular updating (patch management) of operating systems, browsers and office software is the basis. The point that is usually forgotten is the equipment at the edge of the network (edge devices): the firewall, the router, the VPN component and the storage systems. They face the internet, and attackers scan them automatically and systematically as soon as a vulnerability is published.

At the same time, it is worth reducing what is exposed at all (the attack surface). Remote access goes through a VPN or a ZTNA solution with multi-factor authentication, management interfaces are not reachable from the internet, and every opening in the firewall is documented with a reason and a person responsible for it. Whatever is not documented gets closed.

Endpoint and server protection (EDR)

Classic antivirus looks for known files. Modern ransomware changes all the time, and sometimes uses legitimate tools of the operating system itself. An EDR system looks at behavior: a process that starts encrypting many files, an attempt to delete shadow copies, or a suspicious connection to an external server. It can stop the process and isolate the computer from the network (isolation) before the encryption spreads.

The tool alone is not enough. Someone has to receive the alerts and deal with them, at night and on weekends too, because many attacks are timed for exactly those hours. For businesses that have no internal security team, a managed service (MDR) or an external SOC is the way to close the gap.

A backup the attack cannot reach

Backup is the last line of defense, and attackers know it. In many attacks they look for the backup system first, delete or encrypt it, and only then launch the encryption across the network.

A backup that has not been tested is an assumption. A periodic restore test shows how long it really takes to get back to work (RTO), and that is a figure the business’s management should know before the incident, not during it.

Employees and a reporting procedure

Training (security awareness) does not turn employees into security experts, and it does not need to. The goal is for them to recognize a few familiar signs, such as an urgent request to change bank account details (BEC), a sign-in link sent with no context or an unexpected attachment, and to know exactly whom to contact.

The most important thing is that reporting is simple and not embarrassing. An employee who clicked a link and reports a minute later gives you time to stop the incident. An employee who is afraid and stays silent gives the attacker hours. Set one clear reporting channel and respond to every report with thanks, even when it turns out to have been unnecessary.

A plan for the moment of truth (incident response plan)

There is no need for a thick document. One page that everyone knows where to find is enough, also as a printed copy, because on the day of the incident the computers may not work. It should include:

  • Who decides and who talks to whom: management, the IT provider, a legal adviser and the insurance company if there is a cyber policy.
  • First steps: disconnect suspicious computers from the network, and do not turn off or delete anything before consulting, so as not to lose evidence (forensics).
  • Contact details of the external parties, including the National Cyber Directorate’s 119 hotline.
  • The order of recovery: which systems come back first so the business can function again.
  • Who checks whether there is a duty to report to the Privacy Protection Authority.

To make the order of recovery sensible, do a short business impact analysis (BIA) in advance: which systems stop the work when they are unavailable, and how quickly the business must get them back (RTO). These answers turn the response page into a plan that can be carried out, and make it possible to check in advance that the backup actually meets those times.

What to do this week

You do not need to do everything at once. These are steps you can start on right away, in order:

  1. Turn on multi-factor authentication (MFA) for email, remote access and administrator accounts.
  2. Check what is exposed to the internet and close what is not needed.
  3. Make sure there is a separated backup copy and run a restore test.
  4. Separate administrator accounts from working accounts.
  5. Install EDR and decide who handles the alerts.
  6. Write a one-page incident response plan.

Most of these steps are actually carried out by the business’s IT provider. That is why it matters that the provider knows information security in depth, and does not treat it as something to deal with “when there’s time.”