Under the name PHP2Shell, the BSI, the German Federal Office for Information Security, is warning of a vulnerability in the PHP foundation of numerous websites and CMS installations. What is remarkable about this is less the age of the flaw than the fact that it remained without consequence for over a decade: exploitation in practice was so complex that attackers could barely manage it.

It is precisely this hurdle that has now fallen. According to the report, new tools reduce the effort to just a few hours. This fundamentally shifts the risk assessment – a theoretical problem becomes a matter for ongoing operations.

Why old vulnerabilities suddenly become acute

In practice, it is rarely the mere existence of a gap that determines the risk, but the question of how easily it can be exploited. As long as an attack requires deep specialist knowledge, a lot of computing time or manual fine-tuning, it remains the exception. As soon as automated tools take over this effort, the same attack is open to a significantly larger circle.

This shift often catches operators unprepared because the affected component has been in use unchanged for years and has never caused problems. A risk register that was once assessed as "practically not exploitable" and has not been touched since then no longer reflects the current situation.

Who may be affected

PHP is the foundation of a large proportion of content management systems and custom applications operated on the web. The potential circle is correspondingly broad: classic corporate websites, shops, portals, intranets and internal tools built on PHP – regardless of which system runs on top of it.

What matters here is the distinction between application and foundation. An application can be fully maintained and still run on a PHP version that has not received an update for months. Especially in grown environments, responsibility for the application often lies with the agency, for the server with the hosting partner – and a gap in responsibility arises in between.

What operators should check now

Operators should take concrete technical details from the official BSI notices and the publications of the PHP project. Regardless of this, some steps can be taken immediately:

  • Inventory: Which systems run on PHP, in which version, on which server? This also includes secondary areas such as landing pages, staging environments, old campaign pages and deactivated installations that are still reachable.
  • Clarify patch status: Check whether the PHP version in use still receives security updates and whether the current packages have been installed. This applies to operating system packages as well as container images.
  • Record responsibilities: Clarify in writing who updates server packages – the hosting partner, the internal IT team or the agency. Unclear responsibility is the most common cause of delay in such cases.
  • Include deployments: Build pipelines, build images and development environments often use their own PHP versions that differ from the production status. They also belong on the checklist.
  • Review logs: Suspicious access patterns, unexpected files in the web directory or unusual outbound connections are indications that justify closer investigation.

Caution rather than panic

A warning of this kind requires a structured response, not hectic activity. It is helpful to distinguish between urgent and downstream measures: first, publicly reachable, productive systems should be brought to a maintained state. Internal applications and test environments follow afterwards.

Anyone installing updates should test them beforehand in a staging environment – that is, a copy of the live environment that is as identical as possible. A PHP version jump can lead to incompatibilities in older applications. A controlled test costs a few hours; an unprepared update in the production system quickly costs significantly more.

The structural lesson

The case shows a pattern that extends beyond this one gap: the security situation of a system is not a state but a snapshot. It changes not only when new flaws are discovered, but also when known flaws become easier to exploit. Automated tools and machine-assisted analysis accelerate this development.

For operations, this means: the foundation – PHP runtime, operating system packages, database, web server – needs the same maintenance cycle as the application itself. Regular maintenance windows, a documented version status and a clear escalation chain are less spectacular than a fast emergency response, but they prevent exactly that.

For web and software projects, the focus thus shifts from one-time securing to permanent operational capability: anyone who can state the version status of their platform at any time responds to such reports in hours rather than weeks. Plan maintenance and update cycles as a fixed budget item, not as a special case. And contractually establish who is responsible for which layer of the stack – in an emergency, this clarification is often more valuable than any individual technical measure.

Sources