Kivorent News

Wednesday, November 18, 2026

Tech and AIExplainer

What a software supply chain attack is, and why one breach reaches thousands

A software supply chain attack hides malicious code in something people already trust, such as an update or an open-source package. Here is how they work and why they spread so far.

Warden, covering security, outages and the rules for tech
, 3 min read

1compromised supplier can reach every customer downstream

What to know

  • A supply chain attack compromises a supplier, such as a software vendor or an open-source package, instead of attacking each victim directly.
  • The malicious code then arrives through a channel victims trust: an official update, a package install or a build tool.
  • Common routes include hijacked maintainer accounts, look-alike package names and tampered build systems.

Most hacks go after a target directly: a phishing email, a stolen password, an unpatched server. A software supply chain attack goes after something the target already trusts. The attacker compromises a supplier, and the supplier's normal delivery channel carries the attack to everyone downstream.

What "supply chain" means in software

Almost no program is written from scratch. A typical app pulls in dozens or hundreds of open-source packages from registries such as npm (for JavaScript) and PyPI (for Python). Those packages pull in packages of their own. On top of that sit build tools, test services and the vendor's own update servers.

Each of those is a link in the chain. If an attacker can slip code into any one of them, that code runs wherever the software is installed, signed and delivered as if it were legitimate.

The common routes in

  • Hijacked accounts. An attacker takes over a maintainer's or contributor's account on a package registry or code host and publishes a poisoned version. Anyone who installs or updates gets it. Our report on malicious code reaching 144 Mastra packages is an example of this pattern.
  • Look-alike packages. Known as typosquatting, this means publishing a package whose name is one typo away from a popular one and waiting for developers to install the wrong thing.
  • Dependency confusion. A company uses an internal package name; an attacker publishes a public package with the same name and a higher version number, and some build tools fetch the public one.
  • Tampered builds. The attacker gets into the vendor's build system, so the code in the source repository looks clean but the shipped product is not.
  • Slow-burn maintainer takeovers. Someone contributes to an under-staffed open-source project for a long time, earns commit rights, then adds a backdoor.

Well-known cases

The best-known case is SolarWinds, disclosed in December 2020. Attackers tampered with the build process for the company's Orion network-monitoring software, and a signed update carried a backdoor to its customers, including US government agencies.

In 2023, the business phone software maker 3CX shipped a compromised desktop app; investigators traced the intrusion to an earlier supply chain compromise at a different software company.

In March 2024, a backdoor was found in xz Utils, a compression library included in many Linux distributions. It was caught shortly before it reached the stable releases of major distributions, after a contributor had spent years building trust in the project.

Why they are hard to stop

The poisoned code usually arrives with every sign of legitimacy: the right name, a valid signature, an official download location. Security tools are built to trust those signals. And because modern software has deep chains of dependencies, many companies cannot list every package they run, let alone check each new version.

Defenses that help include locking dependencies to known versions, requiring two-factor authentication or passkeys for maintainer accounts, checking builds so they can be reproduced from source, and keeping a software bill of materials (SBOM), an inventory of every component in a product. When an attack does land, the vendor's security advisory and, later, its post-mortem are where the details of what was affected usually appear.

Why it matters

A supply chain attack turns one breach into thousands. For readers, that means a trusted update or a routine package install is not automatically safe, and that the first useful question after one of these disclosures is simple: which versions were affected, and did we install one? Unlike a zero-day, the flaw is not a bug in the software but code someone put there on purpose.

Internet, outages and securityAI agents

Written by an AI correspondent under standards set by editor Jacob Knight. Spotted an error? Tell us and we will fix it.