What a website or server breach is

A breach of a website or server happens when an outsider gains control, full or partial, of a system the organization runs: a company website, an online store, a customer portal, a mail server or a file server reachable from outside. In most cases this is not an attack built specifically against the organization. Automated tools scan the internet constantly looking for systems with a known vulnerability, and a system found to be vulnerable becomes a target.

What happens next varies. Sometimes the website is vandalized and shows a message from the attacker (defacement). More often the attacker prefers not to be noticed: adding code that redirects visitors to other websites, planting code on the payment page that collects credit card details (web skimming), copying the customer database, or using the server to send spam and attack others. A breached internal server can also serve as an entry point into the office network.

What increases the risk

  • Software that was not updated: the content management system (CMS), plugins, themes, code libraries and the server’s operating system. An abandoned plugin that nobody maintains anymore is a risk that only grows over time.
  • Default settings (default configuration): usernames and passwords that were not changed, admin pages at well-known addresses, and folders whose contents can be browsed.
  • Services exposed without need: a database management interface, remote access to the server or the hosting control panel, open to the whole internet and increasing the attack surface.
  • Management interfaces without multi-factor authentication (MFA), and passwords shared among several people or suppliers.
  • Permissions that are too broad: a website running with permission to change any file on the server, or several websites on the same account, so a breach of one opens all of them. The opposite principle is least privilege.
  • No clear ownership: a website built by a supplier who is no longer in the picture, and nobody knows who is supposed to update it.

Early signs

  • Content that changed: pages, links or text that nobody in the organization added.
  • Visitors who report being redirected to another website, mainly from the phone or from search results. Some malicious code works only under certain conditions, so whoever manages the website does not always see it.
  • New administrator users in the website’s system or on the server.
  • Unknown files in the website’s folders, or existing files that changed recently for no reason.
  • Unusual load on the server, or a large amount of outgoing mail the organization did not send.
  • A warning from a search engine or the browser that the website is dangerous, or a notice from the hosting provider.
  • Customers who report unknown charges after buying on the website.

The impact on the organization

A breached website first of all hurts the people who visit it. Visitors who are redirected to fraud sites or receive malware are the organization’s customers, and they will remember where it came from. Search engines and browsers mark such a website as dangerous, and it may drop out of the search results until the problem is handled and verified. The hosting provider may also take the website down to protect other servers.

When the website stores personal information, such as customer details, orders or contact forms, the breach is also an information security incident under the Privacy Protection Law. You need to find out which information was exposed and consider the duty to report to the Privacy Protection Authority. In an online store where payment details were collected, the payment processor and the credit card companies are involved too. And with an internal server, the damage can reach far beyond the server itself, if the attacker used it to get into the network.

How to reduce the risk

  • Patch management at every layer: the operating system, the web server, the programming language, the CMS and the plugins. A component that is not in use is removed completely, because even a disabled component can be vulnerable.
  • Hardening: changing all defaults, limited file permissions, and separation between websites so each one runs in its own account (isolation).
  • Reducing exposure: management interfaces reachable only from authorized addresses or through a VPN, and every opening in the firewall documented with a reason and a person responsible for it.
  • A web application firewall (WAF): a layer that filters suspicious requests before they reach the website.
  • Multi-factor authentication for every management interface: the website, the hosting control panel, the domain registrar and DNS management.
  • Monitoring: tracking changes to files (file integrity monitoring), alerts on new administrator users, and checking the website’s availability.
  • Scans and tests: a periodic vulnerability scan, and a penetration test for systems that hold sensitive information.
  • Ongoing, separated backup: copies of the files and the database, stored outside the server itself (off-site) and tested by restoring from them.

When you discover a breach: the first steps

  • Preserve evidence before changing anything: a copy of the website files, the database and the server logs. Quickly deleting suspicious files also deletes the answer to the question of how the attacker got in.
  • Take the website down temporarily or show a maintenance page, if it is harming visitors.
  • Change all related passwords: the management interface, the hosting, the database, file upload accounts, the domain, and API keys the website uses.
  • Inform the hosting provider, and if the website processes payments, the payment processor as well.
  • If it is an internal server: isolate it from the network and check the computers and servers that were connected to it.

Investigation, cleanup and getting back online with the IR team

An incident response team works in a fixed order. First comes assessment (triage): what was hit, how severe it is, and whether the breach is still active. Then containment: blocking the attacker’s access, including back doors (backdoors) left behind to return. The investigation (forensics) goes through the logs and files to determine the entry point, when the breach began, and which data was exposed.

Based on that date, a clean backup is chosen for the restore, or the server is rebuilt and only checked content is moved to it. Before going back online, the vulnerability is closed, and the system is updated and hardened. Finally comes the stage of reporting and lessons: assessing whether a report to the Privacy Protection Authority is required, asking search engines to review the website again and remove the warning after the cleanup, and changing the process that allowed the breach in the first place. In many cases the most important fix is organizational: deciding who is responsible for updating and monitoring the website from now on.

Summary

  • Most breaches of websites and servers exploit a known vulnerability, a default setting or a service that is exposed without need.
  • Content changes, unusual redirects, new users and unknown outgoing mail are signs that require checking.
  • Updates, hardening, MFA for management interfaces, a WAF, monitoring and separated backup are the basis.
  • When you discover a breach, preserve evidence, change passwords and bring in the IR team before restoring.