CISA has warned businesses that threat actors are actively exploiting a critical vulnerability impacting the password reset function in GitLab.
The vulnerability, tracked as CVE-2023-7028, allows attackers to hijack the password reset process without having to interact with the user. The flaw stems from an improper access control weakness that can be exploited by unauthenticated attackers to force account recovery emails to be delivered to unverified email addresses.
CISA warned that this class of security flaws are “frequent attack vectors” for cyber criminals, due to their simplicity to exploit, and thus pose significant risks to enterprises.
Based on evidence showing the vulnerability was being actively exploited in the wild, CISA added CVE-2023-7028 to its Known Exploited Vulnerabilities (KEV) catalog.
The advisory states federal agencies are legally required to remediate the vulnerability within three weeks, although CISA strongly urged all organizations to reduce their exposure to attacks by remediating the flaw as soon as possible.
GitLab published information on the vulnerability on 11 January 2024, claiming the vulnerability was first introduced on 1 May 2023 after altering their email verification process.
“A change was made in 16.1.0 to allow users to reset their password through a secondary email address. The vulnerability is a result of a bug in the email verification process. The bug has been fixed with this patch, and as mentioned above, we have implemented a number of preventive security measures to protect customers.”
GitLab noted in the January advisory that organizations that have 2FA enabled should be immune to this attack, although it advised users who have received reset emails without triggering them to reset their credentials as soon as possible.
GitLab flaw raises serious industry concerns
This vulnerability is particularly concerning, according to Patrick Wragg, head of incident response at Integrity360, as it targets the mechanism users rely on to re-secure their accounts.
“The GitLab vulnerability is a bad one since the reason you set up an email account is so that you have a safe/secure place to receive password resets,” he said. “Since attackers can just change the email to one under their control, this completely negates this level of security.”
Debrup Ghosh, senior manager at the Synopsys Software Integrity group, said this development should be a reminder to businesses to take active steps to secure their software supply chain.
"CISA's warning about the maximum-severity GitLab vulnerability highlights the complex and urgent need for organizations to understand and proactively manage security risks throughout their software supply chains”, he explained.
Build trust with your employees, customers, and third parties
“More specifically, it highlights that organizations need to look beyond the open source and proprietary code that comprises their applications -- they also need to assess and secure the development tools and platforms they use to build their applications.”
Ghosh outlined how attackers can leverage compromised GitLab accounts to launch further attacks, noting the importance of implementing measures such as multi-factor authentication (MFA) to protect these accounts.
"By compromising vital development tools like code repositories or CI/CD pipelines, which are often trusted inherently and used across multiple applications, malicious actors can launch attacks that are difficult to detect and can have rippling effects downstream,” Ghosh said. “This also demonstrates how important it is to enable two-factor authentication (2FA) for service accounts and other automation bots.”
Wragg argued that this incident shows services that hold highly sensitive information, like GitLab, ought to make MFA mandatory to protect its users.
“The only good part about this vulnerability is that if 2FA is enabled, it won't work. In my opinion, 2FA on a service like Gitlab whereby potentially very sensitive source code is kept for customers should be compulsory.”
Ghosh added that many businesses had failed to patch the issue in their instances of GitLab, demonstrating the lack of robust patch management procedures that are leaving firms vulnerable to attack.
"Additionally, although the patch was issued in January, more than 40% of GitLab instances are not patched almost four months later. Software risk is business risk—hence, organizations need to enable efficient patch management and deployment practices or risk their Intellectual Property being exploited by threat actors."
GitLab's security team told ITPro they are committed to protecting customer data and will continue to work to provide customers with the resources they need to protect themselves in a timely manner.
"Our customers’ security is our top priority, and we are dedicated to ensuring that all aspects of GitLab that host customer data are held to the highest security standards," a spokesperson said.
"GitLab informed customers that a security patch would be released before January 11 to ensure their teams would be available and prepared to address it expeditiously. To the best of our knowledge, the vulnerability was introduced in 16.1.0 on May 1, 2023 and responsibly disclosed through our bug bounty program.
"Self-hosted customers are responsible for the maintenance and security of their systems, including configuration management and deploying patches. GitLab strives to proactively ensure they have access to timely resources to protect themselves. We encourage self-managed customers to review their logs to confirm they were not affected by the vulnerability."
