Russia-linked cyber espionage group Cold Relic, also known as Star Blizzard, compromised a US organization in 2026 by impersonating an employee of another US company it had breached months earlier, according to an investigation by Google’s Mandiant and Google Threat Intelligence Group (GTIG).
The attackers used stolen employee information, corporate contacts and identities obtained during the initial intrusion to conduct targeted phishing against the second organization. They subsequently gained access to its Microsoft Entra ID environment and downloaded files from a compromised user’s cloud storage.
Cold Relic, previously tracked as COLDRIVER and UNC4057, is also known as Callisto, Gossamer Bear and Cold River. The group has operated for more than a decade and has been formally linked by US authorities to Russia’s Federal Security Service (FSB).
According to Google Threat Intelligence Group, the group has compromised government officials, defense contractors and non-governmental organizations across Europe and North America.
Its campaigns focus on obtaining information relevant to Russian government interests, with stolen correspondence and organizational records subsequently used to identify and deceive additional targets.
Cold Relic compromised a US organization’s cloud environment in 2025
The initial intrusion occurred in late 2025, when Cold Relic gained access to the cloud productivity environment of an unnamed US organization.
The attackers used compromised credentials belonging to a Super Admin account and connected through a residential proxy network geographically close to the victim.
They combined the stolen credentials with a multi-factor authentication (MFA) push notification fatigue attack. After the legitimate user accepted an authentication request, the group registered its own device within the organization’s cloud tenant.
The Super Admin privileges allowed the attackers to access administrative controls and extract information across the organization’s cloud environment.
Google reported that Cold Relic downloaded the contents of the organization’s entire cloud storage drive through direct downloads and by sharing large volumes of files with an attacker-controlled email account.
The group also accessed the Admin Console, examined security investigation records and email audit logs to assess the organization’s defensive monitoring capabilities.
Attackers viewed, modified and deleted Alert Center notifications to conceal evidence of their initial access. Using the compromised administrator account, they exported complete employee directories and initiated a domain-wide data export through Google Takeout.
Hackers maintained access using passkeys and rogue MFA devices
Cold Relic maintained access to the first organization for several months through attacker-registered authentication devices, passkeys and MFA backup codes.
The group also synchronized email inboxes with external email clients using OAuth, allowing it to retrieve messages outside the organization’s normal mail interface.
Google found that attacker-enrolled passkeys remained functional after password changes and session resets, enabling the group to authenticate again despite attempts to revoke access.
The attackers searched cloud storage repositories for credentials, personnel records, email archives and other documents relevant to their intelligence collection activities.
Once defenders removed the group from the compromised environment, Cold Relic continued to exploit information obtained during the intrusion.
The attackers created email accounts impersonating employees of the first organization and used those identities to approach internal personnel and external contacts within the same professional networks.
Google said the group used stolen correspondence and relationship information to construct credible messages, increasing the likelihood that recipients would respond to subsequent phishing attempts.
Attackers used stolen identities to compromise Microsoft Entra ID
In the next stage of the campaign, Cold Relic targeted a second US organization that relied on Microsoft Entra ID as its federated single sign-on (SSO) identity provider for Google Workspace.
The attackers impersonated an employee of the previously compromised organization and established communications with their new target through a secure messaging application.
They presented themselves as a legitimate professional contact and shared phishing links under the pretext of providing a document concerning a technical issue. The links were designed to collect account credentials and Entra ID authentication tokens. After obtaining valid credentials, Cold Relic authenticated to the second organization’s environment and registered a new MFA device within its SSO system. The additional device allowed the attackers to retain access to the compromised account during the intrusion. Google confirmed that the group downloaded files in bulk from the affected user’s cloud drive.
The attack differed from an isolated phishing campaign because the impersonation relied on information collected during a previous organizational compromise. The attackers already possessed identities and professional relationship details that could make their communications appear legitimate. Google assessed that this familiarity increased the probability of a successful phishing attempt.
Following the second compromise, the attackers changed the profile name associated with the secure messaging account to another identity. Google believes the revised persona may have been intended for further attacks against organizations in the same professional network.
Google identifies persistent cloud identity abuse by Russian hackers
Google’s investigation found that Cold Relic combined social engineering with the abuse of legitimate cloud authentication mechanisms to maintain access and collect information across multiple organizations.
The group registered unauthorized MFA devices, created passkeys and retained backup authentication codes while conducting extensive searches of email and cloud storage repositories. Its activity included collecting operational documents and employee information that could support subsequent attacks against business contacts and professional associates.
The use of compromised cloud administrator privileges also gave the attackers access to security logs and alert management functions, which they manipulated during the initial intrusion.
Google’s findings indicate that password resets and session revocation did not immediately eliminate the group’s access when attacker-controlled passkeys remained registered.
The investigation also documented the use of OAuth-based email synchronization to extract messages through external clients. These techniques enabled Cold Relic to continue collecting data after its initial access and to use information from one victim against another.
Google attributed the campaign to an intelligence collection strategy in which compromised professional relationships become a means of reaching additional organizations.
The report did not identify either US victim organization or disclose the total volume of information stolen during the two intrusions.
The group has maintained a consistent operational tempo against high-value targets, according to GTIG, using access to corporate directories, correspondence and cloud documents to support further credential theft and targeted compromise.
Click Here For The Original Source.
——————————————————–
