Written by: Mio Kanzaki(Security Analyst (AI))Reviewed by: 千葉圭介Updated
Ransomware prevention means reducing, at the entry points, attacks that encrypt or steal your data and demand money, and being able to keep the business running if one happens. It concerns companies that use an internal network or a VPN (a way to connect safely to the company network from outside), companies with their own servers or member sites, and the agencies and IT firms that maintain them. Three jobs come first: inspect VPN devices and remote connections, strengthen logins, and confirm that a backup stored apart from the network can be restored.
This page is a hub for executives and IT staff. If a ransom demand is already on a screen, read the emergency guide first.
What does a ransomware attack do?
The name means software that demands a ransom. Damage does not stop at files that will not open. Attackers first get inside, take their time to carry important data out, and only then encrypt terminals and servers all at once. By the time encryption is found, the data has often already been stolen.
The demand has two layers: "pay if you want your data back" and "pay or we publish what we stole". The second is called double extortion. The National Police Agency also confirmed 17 cases in 2025 in which criminals demanded money only because they had stolen data, without encrypting anything. This is called no-ware ransom.
So think of the damage in three kinds: business stops, data is stolen and published, and harm spreads to business partners and customers. A backup helps with the first kind only. Closing the entry points and preparing to restore are separate jobs.
Should you pay the ransom?
Nobody can say that paying brings the data back. There are accounts of data that could not be decrypted after payment, of another demand arriving after payment, and of data being published. If you do not pay, the business may stay stopped, so this is a heavy decision.
Before deciding, find out how far the damage reaches, consult the police and judge whether a legal reporting duty applies. We do not negotiate with criminals or pay ransoms on anyone's behalf. We handle recovery work and a review of the settings that prevent a repeat.
Where do attackers get in?
Knowing the entry point decides the order of repair. The table lists the main entry points that lead to ransomware damage, arranged so that a small or medium-sized company can check them.
| Entry point | What happens | What to check | Read more |
|---|---|---|---|
| A hole in a VPN device | An attacker enters the network through a device whose fix was not released or not applied | The firmware version of the device and the settings exposed to outside | VPN appliance vulnerabilities |
| Exposed remote desktop (remote screen sharing) | Login is tried again and again against a terminal reachable directly from the internet | The list of terminals reachable from outside, and counts of failed logins | Exposed remote desktop (RDP) |
| Reused credentials, no multi-factor authentication | A leaked ID and password let someone in as a legitimate user | Whether two-factor authentication is on, and accounts of people who left or of contractors | Missing multi-factor authentication |
| Software that was not updated | A published hole is abused and the system is controlled remotely | Update state of the OS, server software and CMS | Unpatched software |
| Permissions of partners and contractors | An attacker uses the permissions of a maintenance company as a foothold | The connection routes and permissions you gave contractors | Preventing supply chain attacks |
| Impersonating emails and calls | A convincing message makes you type your own credentials | Whether there is a rule to report suspicious contact | Preventing phishing |
After getting in, an attacker can widen the reach in steps, from a terminal with weak rights to administrator rights (privilege escalation). Just before encryption, backups are sometimes deleted or encrypted too, so fix backup failures beforehand.
We do not publish the steps or programs that attackers actually use.
What do public statistics show?
Public figures show how common ransomware is and how it gets in. The numbers below were taken from the original publications.
National Police Agency: more than 60% of victims are small and medium-sized companies, two-thirds of entries through VPN devices
The National Police Agency counted 226 ransomware cases reported to police in 2025 (Reiwa 7): 116 in the first half and 110 in the second. By company size, 64 were large companies and 143 were small and medium-sized companies, which is more than 60% of the total (2026 White Paper on Police).
The routes of entry come from a survey of victim companies with 92 valid answers. 61 (66.3%) were entered through a VPN device and 19 (20.7%) through remote desktop. Together that is 80 of the 92 answers.
Among the 153 cases in which money was demanded and police could confirm the method, double extortion, which threatens to publish stolen data, accounted for 136 (88.9%). Even when a backup restores operations, the problem of stolen data remains.
IPA: ransomware has been selected for 11 years in a row and ranks first in 2026
IPA issued its Top 10 Information Security Threats 2026 on 29 January 2026. First among threats to organizations was damage from ransomware attacks, which has been selected for 11 consecutive years. Second was attacks aimed at the supply chain or contractors, third was cyber risk from the use of AI (new this year), and fourth was attacks that exploit system vulnerabilities.
Attackers use AI as well
The security company Semperis states in its 2026 Ransomware Risk Report (October 2026) that AI was used in some form in 84% of the attacks.
After reading the statistics, you may want to see how your entry points look. The free URL check reads only what anyone can see from outside, and it costs JPY 0.
What do you do in the first hours after an attack?
Hesitating over decisions widens the damage, so here is how to act when it happens.
- Disconnect the terminals and servers you suspect from the network. Pull the LAN cable or turn off Wi-Fi. Whether to power off can destroy evidence, so decide after consulting a specialist.
- Photograph the screen and the ransom note. Do not delete or reset files on the terminals.
- Consult the police (the cybercrime consultation desk of the prefectural police). A report of the damage is also used with business partners and insurers.
- If personal data was included or may have been, you may have to report to the Personal Information Protection Commission and notify the people concerned. The next section sets this out.
Does a reporting duty under the personal information law apply?
When personal data may have leaked, the Personal Information Protection Commission describes four kinds of situation that require a report. They are: data includes special care-required personal information such as medical history, there is a risk of financial harm, the act was done with an unlawful purpose such as unauthorized access, or the number of people concerned exceeds 1,000. Ransomware intrusion can count as unauthorized access, so it can fall into one of these.
A report has two stages. The Commission asks for a prompt preliminary report as soon as the leak is found. Its guide gives 3 to 5 days from the day of discovery as a guide. The final report is due within 30 days of discovery, or 60 days where an unlawful purpose is suspected. Whether you are covered depends on the case. When in doubt, read the Commission's guide or consult a specialist.
What should you do first, by company size?
Start with the steps that suit the shape of your company.
A company of a few people up to about 20, with site and devices outsourced
With nobody in-house who knows the field, look first for devices that nobody updates. Work in this order: list the devices and services exposed to outside, name who updates the VPN device, add two-factor authentication and store backups apart from the network.
Ask the company that maintains your systems, in writing, who updates the firmware of the VPN device and router, and when. If the answer is vague, treat the devices as not updated until you confirm otherwise.
A company of a few dozen with several offices or partner connections
The number of entry points grows here. VPN devices were two-thirds of the entries in the police survey, so inspect VPN appliances and remote desktop first.
Also take stock of the connection routes and permissions you have given partners and maintenance companies. Accounts of people who left, or of firms whose contract ended, may remain. Consider dividing the network by department so that one infected machine does not spread to the rest (network segmentation).
A company with a member site or its own system that holds data
Stolen data alone can bring a reporting duty or the need to explain to partners, even if nothing is encrypted. Both the site's servers and the internal network can be entry points.
Build four things into operations: an assessment before launch and after each large change, a narrowed entrance to the admin screen, detection of rewrites, and an emergency contact decided in advance. If you use the cloud, review permissions and public settings as well (cloud settings assessment).
10 things to do today
These 10 are an order for reducing entry points on your site and internal systems and for building a way back. Several follow our fix menu, and item 10 is a rule for how you operate. Each item notes whether the admin screen is enough or a specialist is the safer choice.
- Update everything exposed, and delete parts you do not use. Bring VPN devices, routers, server operating systems, and the WordPress core, themes and plugins up to date. Remove what you do not use. For a device update, pick a quiet time and keep a copy of the settings first.
- Tidy administrator accounts and passwords. Remove administrators who are no longer needed and end reused and weak passwords. Check for accounts of people who left and of contractors whose contract ended.
- Turn on two-factor authentication. If a password leaks, the account stays shut without the phone confirmation. Start with the VPN, email and admin screens (two-factor authentication).
- Restrict where admin screens and remote operation can be opened from. Allow only fixed places or terminals. Do not expose remote desktop directly to the internet.
- Close entrances you do not use. Shut off ports and channels you never use.
- Take daily backups and keep them apart from the network. Make three things one set: take it, disconnect it, and test a restore. A backup never restored may not come back.
- Detect file tampering. Set up something that notices changes such as suspicious files appearing while the look is unchanged, or an administrator being added without your knowledge.
- Add a WAF, which blocks harmful traffic at the door. Dangerous requests are dropped before they reach the site, which also buys time until a fix is available (WAF).
- Put important files out of reach and tighten permissions. Put configuration files and customer data where they cannot be read from outside, and stop using administrator rights for daily work.
- Put contacts and steps for an attack on one sheet. Write the police desk, your maintenance company, whom to consult including us, and who tells business partners on one page, and print it. When the network is down, a file on a screen cannot be opened.
Start with items 1 to 3. From item 4 onward, a wrong setting can disrupt business, so take a backup first. All 10 done is still not a guarantee against an attack. The work lowers the odds and limits the harm if one comes.
Not enough hands in-house? Ask us which of the 10 your company has not yet covered.
What should you hand to a specialist?
Before you decide who does a job, ask whether a mistake can be undone and whether the job needs skill to examine devices and servers.
We recommend a specialist for these:
- Assessment of entry points visible from outside. The version of a VPN device and ports left open can be checked from outside. Our server and network assessment is from JPY 49,800 (excl. tax, up to 5 IPs), and one re-assessment after fixes is free. The price rises with the number of IPs.
- WordPress/CMS site assessment. Holes in a site cannot be seen by looking at it. The assessment is from JPY 49,800 (excl. tax, up to 30 pages). The amount moves with screens behind login and the number of paid plugins.
- Investigation and recovery when you suspect damage. Evidence can disappear if files are touched first, so the order of steps matters. Recovery, including removing harmful code and setting up recurrence prevention, is from JPY 29,800 (excl. tax). A same-day start adds JPY 20,000 (excl. tax). The amount rises with the number of infected devices and how far the damage spread, and a formal quotation follows after we hear the situation.
- Continuous watch. Monthly updates, backups and tamper monitoring lapse as soon as someone forgets. Maintenance Lite is JPY 9,800 per month (excl. tax).
- Answers to business partners' check sheets, and internal policies. The security advisory plan is JPY 29,800 per month (excl. tax). It includes unlimited consultation and one online meeting a month.
We assess the entry points that are visible from outside, confirm and fix holes we find, investigate and recover after damage, and keep watch afterwards. A PDF of changes comes before delivery, and you confirm the actual screens. Support is free for one month after delivery.
We should also say what we do not do. Running anti-virus software on office computers is not our main target. Guard, our AI monitoring for WordPress (Lite JPY 5,000 per month and Pro JPY 15,000 per month, both excl. tax), watches a website. It is not ransomware protection for a whole internal network. We cannot promise protection, because attack methods keep changing. An assessment checks the holes known now and does not secure the future.
Which pages cover the individual holes?
Holes tied to ransomware have their own pages, covering how they work, how to check and how to fix.
Entry points
Why damage spreads after entry, and how to prepare
The other hubs are hacking prevention, preventing unauthorized access and preventing information leaks.
How can the free URL check help you decide?
Know your position before deciding what to fix. Put the address of your site into the free URL check. It reads public information only, such as the expiry of the encryption certificate, security headers, versions of public components with their known holes, and email authentication settings. It makes no login attempts and sends no attacks. It costs JPY 0.
If the result worries you, we sign an NDA, obtain your assessment consent form and run the free assessment, in which tools and a person make a quick check. The issue report reaches you before any decision on fixes. A formal quotation follows.
Frequently asked questions
If we pay the ransom, will we get our data back?
There is no guarantee. There are cases where data was not decrypted after payment, and cases where stolen data was published. First stop the damage from spreading, then consult the police and a specialist. We do not negotiate with criminals or pay ransoms on anyone's behalf.
Can a small company be hit by ransomware?
Yes. Police data for 2025 show 143 of 226 reported cases at small and medium-sized companies. In the police survey, VPN devices were the most common entry point, so keeping them updated matters.
Is having backups enough for ransomware prevention?
No. Backups themselves are sometimes encrypted, and some companies hold backups they cannot restore. You need storage disconnected from the network and a test that a restore works. Stolen data does not come back, so closing the entry points is needed in parallel.
How much does ransomware prevention cost?
The free URL check, which looks only at what is visible from outside, costs JPY 0. A server and network assessment, which covers VPN devices and open ports, is from JPY 49,800 (excl. tax, up to 5 IPs). The amount moves with the number of IPs. A formal quotation follows the free assessment.
Next step
Today, start on items 1 to 3 above, and look at your own site with the free URL check.
To settle the order of work and what to outsource, ask us. If you already suspect an attack, start from the emergency guide.
Written by Mio Kanzaki (AI security analyst). Supervised by 千葉圭介. Updated 19 November 2026.
Frequently asked questions
If we pay the ransom, will we get our data back?
There is no guarantee. There are cases where data was not decrypted after payment, and cases where stolen data was published. First stop the damage from spreading, then consult the police and a specialist. We do not negotiate with criminals or pay ransoms on anyone's behalf.
Can a small company be hit by ransomware?
Yes. Police data for 2025 show 143 of 226 reported cases at small and medium-sized companies. In the police survey, VPN devices were the most common entry point, so keeping them updated matters.
Is having backups enough for ransomware prevention?
No. Backups themselves are sometimes encrypted, and some companies hold backups they cannot restore. You need storage disconnected from the network and a test that a restore works. Stolen data does not come back, so closing the entry points is needed in parallel.
How much does ransomware prevention cost?
The free URL check, which looks only at what is visible from outside, costs JPY 0. A server and network assessment, which covers VPN devices and open ports, is from JPY 49,800 (excl. tax, up to 5 IPs). The amount moves with the number of IPs. A formal quotation follows the free assessment.