The disclosure of CVE-2026-105445, a critical vulnerability in Docker Hub, highlights a serious weakness in software supply chain security: an attacker does not necessarily need to compromise a developer's password or exploit an application vulnerability to distribute malicious software through a trusted repository.

The vulnerability existed in Docker Hub's identity and access management process when converting personal accounts into organizations. During this conversion, personal access tokens (PATs) were reassigned to the newly designated organization owner without rotating their secrets or obtaining the new owner's consent.

This created a dangerous situation. An attacker could create a Docker Hub account, generate an access token and convert that account into an organization while naming another user as its owner. The token would then be associated with the targeted user's identity, even though the attacker still possessed the original secret.

Depending on the token's permissions and the targeted user's access rights, the attacker could potentially download, publish or delete container images in repositories belonging to that user or organizations they could access.

The most concerning implication is the possibility of software supply chain compromise. By publishing a malicious container image under a legitimate repository name, an attacker could potentially distribute compromised software to organizations that trust that publisher. Automated CI/CD pipelines and container deployment systems may subsequently retrieve and deploy the modified image, allowing malicious code to enter development, testing or production environments.

Unlike conventional malware distribution, such an attack could exploit an established relationship of trust. The container image might appear to originate from a recognized developer or organization, while its contents had been modified without authorization.

This demonstrates why software supply chain security cannot rely solely on repository names, publisher identities or the apparent legitimacy of software distribution platforms.

Organizations must strengthen credential lifecycle management, enforce least-privilege permissions for access tokens, use dedicated organizational credentials for automated workflows and continuously monitor unusual image publishing activities. Container image signing, signature verification, immutable image digests, vulnerability scanning and runtime monitoring provide additional layers of protection.

The incident also highlights the importance of treating identity changes as security-sensitive operations. Credentials should never silently acquire another user's privileges simply because ownership information changes within an application.

Docker has confirmed that the vulnerability was fixed on September 30, 2026, and that its investigation found no evidence of exploitation after reviewing approximately 304,000 image pushes involving transferred tokens. This is reassuring, although the underlying design weakness remains an important lesson for the wider industry.

The broader cybersecurity concern is not simply that attackers could impersonate a Docker Hub user. It is that they could potentially exploit that identity to distribute malicious software through repositories that developers and automated deployment systems already trust.

Software supply chain security must establish not only where software comes from, but also whether its publisher, contents and integrity can be independently verified before deployment.

Docker has disclosed a critical vulnerability (CVE-2026-105445) in Docker Hub that let attackers access another user's container repositories using a personal access token (PAT). [...]

Source: Docker Hub flaw let attackers impersonate users, tamper with images via Bleeping Computer — published 13 Oct 2026.