Critical Docker Hub Flaw Lets Attackers Impersonate Users, Push Images

Key Takeaways

• A critical vulnerability (CVE-2026-105445) in Docker Hub allowed attackers to impersonate users and push malicious container images.
• The flaw stemmed from improper handling of Personal Access Tokens (PATs) during Docker Hub account conversions.
• While the vulnerability posed a significant software supply chain risk, Docker found no evidence of exploitation.
• Docker has patched the issue, revoked affected tokens, and implemented new security measures.

Docker has revealed a critical security vulnerability within its Docker Hub platform that could have permitted malicious actors to impersonate legitimate users and publish unauthorized container images under trusted identities. Identified as CVE-2026-105445, this flaw specifically impacted personal access tokens (PATs) during the process of converting individual accounts into organizational structures.

The company addressed the issue on September 30, 2026, and its subsequent investigation concluded that there was no evidence of this vulnerability being exploited in the wild. Despite the absence of confirmed attacks, the flaw presented a substantial software supply chain risk, as attackers could potentially have compromised or altered container images that developers widely rely upon for their applications and systems.

Depending on the specific permissions granted to a compromised token and the access level of the targeted user, an attacker could have performed operations such as pulling, pushing, or deleting images within personal repositories and associated organizations. It is important to note, however, that the vulnerability did not facilitate a complete account takeover, meaning attackers could not gain full control over a victim’s Docker Hub account beyond image registry operations.

Docker Hub Vulnerability Details

How the Flaw Operated

Docker Hub offers users the convenience of converting a personal account into an organization account, a feature designed to help teams maintain consistent image references for their build systems and applications. This process typically involves selecting an existing Docker Hub account to assume ownership of the newly formed organization.

Starting from August 26, 2024, a critical oversight in this conversion process caused the original account’s personal access tokens (PATs) to be transferred to the designated new owner. Crucially, Docker’s system updated the ownership record for these tokens without either modifying the token secret itself or revoking its validity. Furthermore, the new owner was neither prompted to approve this transfer nor informed that it had occurred.

This mechanism created a dangerous scenario: an attacker could create their own Docker Hub account, generate a PAT, and then convert that account into an organization, deliberately naming a victim as its owner. The attacker would retain knowledge of the original token secret, but Docker Hub would now associate that token with the victim’s identity. Consequently, any registry requests made using this specific token would authenticate as the targeted user.

The vulnerability represented a failure in how credential ownership was managed, rather than a direct theft of passwords. The affected tokens remained functional even after their associated identity changed. Docker’s investigation determined that the exposure window for this flaw spanned from August 26, 2024, at 20:33 UTC, until the production fix was deployed on September 30, 2026, at approximately 19:10 UTC.

The ability for an attacker to publish an image under a legitimate repository name could have exposed downstream users to unexpected and potentially malicious code, undermining the integrity of software dependencies.

While this incident is distinct from past supply chain attacks, such as the Trivy supply chain compromise involving Docker images, it underscores the critical importance of secure publishing pathways for container images. Docker has confirmed that there is no evidence this specific vulnerability was exploited by malicious actors.

Scope of Impact

The affected tokens did not grant access to the Docker Hub website interface itself. Attackers could not leverage this flaw to alter a victim’s password, email address, or two-factor authentication settings. The impact was confined to the registry level, with the extent of potential damage depending on the scope of the individual token’s permissions and the target user’s repository access rights. The issue was responsibly disclosed to Docker through its vulnerability disclosure program by cybersecurity researchers Peter Molnar and Krisztian Hozak from mptrdev.com.

According to Docker’s security advisory, the company was able to reproduce the vulnerability using internal test accounts and deployed a fix within four hours of receiving the initial report.

Post-Discovery Actions and Mitigation

Following the discovery, Docker conducted a comprehensive review of all account conversions that had occurred since the vulnerability’s introduction. This included scrutinizing registry authentication logs, usage data, and token records. Investigators examined approximately 304,000 image pushes executed with transferred tokens. Their analysis revealed that all these pushes were directed at repositories belonging to the converted organization or the new owner, with no instances observed reaching unrelated accounts. Docker also inspected higher-risk public images, without executing them, and found no evidence of malware, cryptominers, credential theft, persistence mechanisms, or tampering.

However, registry logs only record token usage, not the mere existence of a token. Consequently, Docker took a proactive stance, treating every transferred PAT as potentially compromised, including those that never appeared in usage logs.

Docker initiated notifications to all potentially exposed users and organization owners on October 2. By October 8, the company deactivated all transferred tokens, with the exception of those that owners explicitly requested to retain. The account conversion process has since been updated to no longer transfer PATs. Docker is also implementing additional security enhancements, including owner consent mechanisms, notifications, and improved record-keeping for token ownership changes. Users who did not receive a notification from Docker were not exposed to this vulnerability and do not need to take any incident-specific actions.

What You Should Do

• Regularly review your Docker Hub Personal Access Tokens (PATs) and revoke any that are unfamiliar, unused, or have excessive permissions.
• For organizational automation and integrations, prioritize the use of organization-specific access tokens instead of personal ones.
• Implement robust container image scanning tools as part of your CI/CD pipeline to continuously monitor the security of images you use and publish.
• Stay informed about Docker’s security advisories and promptly apply any recommended updates or security measures.

Disclaimer: HackersRadar reports on cybersecurity threats and incidents for informational and awareness purposes only. We do not engage in hacking activities, data exfiltration, or the hosting or distribution of stolen or leaked information. All content is based on publicly available sources.