Misconfiguration in Zero-Trust Solution Could Allow Threat Actors to Bypass 2FA
The Adlumin team recently investigated a security incident in which a malicious actor(s) successfully managed to gain unauthorized access to a company’s networks by completely bypassing Duo, a popular zero-trust security solution used by hundreds of organizations worldwide.
Background
The incident occurred in early February 2024 when threat actor(s) used two compromised sets of email credentials to log in remotely to the targeted company’s network from servers with IP addresses registered to Russia and Brazil. Subsequently, the company’s security tools, including Adlumin, generated several alerts for malicious activity detected within the network. This activity included credential brute forcing attempts, attacks against Microsoft Active Directory and Kerberos, and the use of Netscan to enumerate endpoints and servers.
Security teams responded to the alerts and successfully halted and locked out the threat actors before they could inflict more harm on the network, but questions remained as to why Duo’s two-factor authentication (2FA) was not prompted to verify the legitimacy of the login sessions which would have protected against compromised credential-based attacks.
Investigation Findings
The Adlumin investigation revealed that the two compromised email accounts used by the threat actor(s) were stale accounts which had been mistakenly configured with a policy that allows for unenrolled or partially enrolled users to authenticate into their network without 2FA.
According to Duo’s online documentation (last updated on Jan. 29, 2024), a “New User Policy” to allow access without 2FA, does not prompt users to complete enrollment and they are granted access without two-factor authentication.1
This type of user policy is made available to organizations for several reasons, including facilitating a gradual rollout of 2FA within the organization or a slow adoption of new zero-trust practices. However, it remains important to monitor events generated by users that bypass 2FA. Duo does offer such a monitoring feature to companies using Duo Premier, Duo Advantage, and Duo Essentials Plan.
With any 2FA solution, it’s important to consider the risks of enabling or using user policies that bypass it in any scope. Bypassing 2FA for certain users or scenarios reduces the overall security posture of the system and network. It can create fringe but exploitable instances where authentication relies solely on a single factor (e.g., username and password) that may be more susceptible to compromise – which was the case in the security incident investigated by Adlumin.
When users are not required to use 2FA, there is an increased vulnerability window. Attackers may exploit this period, especially if users with reduced authentication factors can enable access to sensitive information or critical systems.
In its online documentation, Duo does warn account owners and administrators who configure login access to remember that users with bypass status are not subject to restrictions and can bypass Duo authentication entirely.2
Conclusion
To protect against similar attacks at organizations that use Duo or other zero-trust solutions, Adlumin recommends that companies and organizations ensure user access policies are correctly configured and consider the security risks that come with allowing some users to bypass 2FA.
Organizations can avoid or reduce their exposure to an attack by practicing good account hygiene. This includes routinely conducting account reviews to identify and deactivate accounts that are no longer needed, establishing efficient communication between IT departments and human resources when employees leave an organization, and automating account provisioning and deprovisioning processes.
Indicators of Compromise (IOCs)
References:
[1] https://duo.com/docs/policy#new-user-policy
[2] https://duo.com/docs/policy#overview
