COLD RELIC Hackers Pivot Through Google Workspace to Breach Second US Org

Russian COLD RELIC Hackers Use Stolen Google Workspace Data to Breach Second U.S. Organization

Russian threat actor COLD RELIC successfully compromised two U.S. organizations through a sophisticated multi-stage attack. After initially breaching the first victim’s Google Workspace environment, the attackers extracted credentials and business relationship data to pivot and compromise a second organization. This attack highlights how cloud platform compromises can serve as launching pads for supply chain attacks, emphasizing the critical need for comprehensive cloud security monitoring, conditional access policies, and incident response procedures that account for lateral movement beyond organizational boundaries.

Introduction

Cloud platforms have become the modern enterprise’s backbone, but they’ve also emerged as high-value targets for sophisticated threat actors. A recently disclosed campaign by the Russian threat group COLD RELIC demonstrates how attackers are weaponizing compromised cloud environments to launch cascading attacks across organizational boundaries.

The attack chain began with a successful breach of one U.S. organization’s Google Workspace environment. Rather than stopping there, COLD RELIC operators mined the compromised account for valuable intelligence—specifically targeting business communications, vendor relationships, and stored credentials. This reconnaissance enabled them to identify and subsequently compromise a second U.S. organization with which the first victim maintained business relationships.

This attack pattern represents a concerning evolution in threat actor tradecraft, where cloud compromises serve not as endpoints but as pivot points for broader campaigns targeting interconnected business ecosystems.

Background & Context

COLD RELIC operates within the broader landscape of Russian cyber espionage groups, though intelligence on this specific actor remains limited compared to more established groups like APT28 or APT29. The group’s focus on Google Workspace environments aligns with broader targeting trends, as cloud productivity platforms have become repositories for sensitive business intelligence, credentials, and relationship data.

Google Workspace environments present attractive targets for several reasons. First, they contain email communications that reveal business relationships, operational details, and sometimes embedded credentials. Second, they often store documents with sensitive technical information, network diagrams, or authentication details. Third, compromised workspace accounts can provide legitimate-appearing access that bypasses traditional perimeter defenses.

The supply chain attack vector demonstrated in this campaign reflects a methodology increasingly favored by nation-state actors. Rather than directly targeting hardened networks, attackers compromise trusted partners or service providers to gain indirect access to ultimate targets. This approach exploits the trust relationships inherent in modern business ecosystems.

Technical Breakdown

The COLD RELIC attack chain unfolded across multiple stages, each building on intelligence gathered from previous compromises.

Initial Access and Workspace Compromise

The attackers first gained unauthorized access to Organization A’s Google Workspace environment. While the initial access vector hasn’t been publicly disclosed, common techniques include credential phishing, session hijacking, or exploitation of weak authentication controls. Once inside, the attackers established persistent access to email accounts and cloud storage.

Reconnaissance and Data Mining

With workspace access established, COLD RELIC operators conducted extensive reconnaissance. They specifically targeted:

  • Email communications between Organization A and its business partners
  • Stored credentials in emails, documents, or spreadsheets
  • Vendor relationship documentation
  • Project collaboration details that revealed technical architectures
  • Contact information for employees at partner organizations

This intelligence gathering phase is crucial to understanding the attack’s sophistication. Rather than immediately exfiltrating data or deploying ransomware, the attackers invested time in understanding the victim’s business ecosystem to identify secondary targets.

Credential Extraction and Validation

The attackers extracted credentials found within the compromised workspace. These may have included:

  • Shared service account credentials
  • API keys for third-party integrations
  • VPN credentials shared via email
  • Application passwords stored in documents

The attackers then validated these credentials against Organization B’s systems, identifying which stolen credentials still functioned and provided meaningful access.

Secondary Compromise

Using the harvested credentials and intelligence about the business relationship between organizations A and B, COLD RELIC successfully authenticated to Organization B’s systems. The legitimate business relationship provided cover for the malicious access, as connections between the two organizations wouldn’t immediately trigger security alerts.

Impact & Risk Assessment

This attack demonstrates several critical risks facing organizations that rely on cloud platforms and maintain interconnected business relationships.

Direct Impact

Both compromised organizations face immediate consequences including unauthorized access to sensitive data, potential intellectual property theft, and exposure of customer information. The attackers maintained persistent access to email communications, potentially compromising confidential business strategies, financial information, and operational details.

Cascading Risk

The more concerning impact involves cascading risk across the business ecosystem. When Organization A was compromised, it effectively became a liability to every partner with which it maintained trust relationships. Organizations B, C, D, and beyond all became potential targets simply by virtue of their business connections.

Detection Challenges

This attack pattern presents significant detection challenges. Access from Organization A to Organization B may appear completely legitimate, as these organizations already maintained authorized data-sharing or collaboration relationships. Traditional security controls that whitelist partner organizations or trusted IP ranges may actually facilitate attacker access.

Long-Term Consequences

Organizations may face regulatory penalties if the compromise resulted in data protection violations, particularly if customer or employee data was exposed. Additionally, the breach erodes trust between business partners, potentially damaging commercial relationships that took years to establish.

Vendor Response

Google has not publicly disclosed specific details about its response to this campaign, though the company maintains several security features designed to prevent and detect such compromises.

Google Workspace includes security capabilities such as:

  • Advanced protection program for high-risk users
  • Context-aware access controls based on device, location, and user identity
  • Anomaly detection for unusual account behavior
  • Security investigation tools for administrators
  • Mandatory two-factor authentication capabilities

Organizations using Google Workspace are encouraged to review their security settings and enable advanced protection features. Google regularly publishes security advisories and best practices through its Cloud Security blog and workspace administrator resources.

The broader security community has emphasized the need for enhanced monitoring capabilities that can detect when compromised credentials are used from unexpected contexts, even when the access appears technically legitimate.

Mitigations & Workarounds

Organizations can implement several controls to reduce exposure to this attack pattern.

Strengthen Authentication Controls

Implement phishing-resistant multi-factor authentication across all cloud platforms:

gcloud identity groups memberships modify \
  --organization=ORGANIZATION_ID \
  --require-security-key

Implement Conditional Access Policies

Configure context-aware access controls that evaluate device health, location, and behavioral patterns:

# Example conditional access policy structure
ConditionalAccessPolicy:
  conditions:
    - user_risk: high
    - location: unexpected_geography
    - device_compliance: false
  action: block_or_require_mfa

Credential Hygiene

Never store credentials in emails or documents. Use dedicated secrets management solutions instead:

# Use secrets managers for credential storage
gcloud secrets create api-key \
  --data-file=/path/to/secret \
  --replication-policy="automatic"

Network Segmentation

Implement zero-trust architectures that don’t automatically grant access based on partner relationships. Every access request should be authenticated and authorized independently.

Detection & Monitoring

Security teams should implement monitoring strategies specifically designed to detect this attack pattern.

Anomalous Access Patterns

Monitor for workspace access from unusual locations, devices, or times:

# Example: Monitor for access from unexpected geographic locations
def detect_anomalous_access(user_logs):
    baseline_locations = get_user_baseline_locations(user_id)
    for log in user_logs:
        if log.location not in baseline_locations:
            alert(f"Anomalous access from {log.location}")

Credential Usage Monitoring

Track when stored credentials are accessed within workspace environments:

-- Example: Query workspace audit logs for document access
SELECT user_email, doc_title, access_time, ip_address
FROM workspace_audit_logs
WHERE doc_title LIKE '%credential%' OR doc_title LIKE '%password%'
  AND access_type = 'download'
ORDER BY access_time DESC;

Cross-Organization Access Analysis

Monitor for unusual patterns in partner organization access:

# Example: Analyze authentication logs for partner access patterns
grep "partner_org_domain" /var/log/auth.log | \
  awk '{print $1, $2, $3, $9}' | \
  sort | uniq -c | sort -rn

Data Exfiltration Indicators

Watch for bulk downloads or unusual data transfer patterns:

# Example: Detect bulk email exports
def detect_bulk_exports(workspace_logs):
    for user in workspace_logs:
        if user.export_count > THRESHOLD:
            alert(f"{user.email} performed {user.export_count} exports")

Best Practices

Organizations should adopt comprehensive security practices that address both technical and procedural aspects of cloud security.

Implement Zero-Trust Architecture

Adopt a zero-trust model where no user or system is automatically trusted based on network location or business relationships. Every access request requires authentication, authorization, and continuous validation.

Regular Security Audits

Conduct quarterly reviews of workspace permissions, shared documents, and authentication logs. Identify and revoke unnecessary access privileges, remove stale accounts, and validate that security controls function as intended.

Incident Response Planning

Develop incident response procedures that explicitly address cloud compromises and their potential for enabling lateral movement to partner organizations:

  • Define escalation procedures when cloud compromise is detected
  • Establish communication protocols with business partners
  • Create runbooks for rapid credential rotation
  • Document procedures for notifying potentially affected third parties

Security Awareness Training

Educate employees about the risks of storing credentials in emails or documents. Emphasize the importance of reporting suspicious access or unusual account behavior.

Vendor Risk Management

Establish security requirements for business partners and service providers. Conduct regular security assessments of organizations with privileged access to your environment.

Implement Privileged Access Management

Use dedicated privileged access management (PAM) solutions for administrative accounts. Never use standard workspace accounts for privileged operations.

Key Takeaways

The COLD RELIC campaign demonstrates several critical lessons for security practitioners:

  • Cloud compromises enable supply chain attacks: A breach of one organization’s cloud environment can serve as a launching pad for attacks against business partners, vendors, and customers.
  • Trust relationships create security dependencies: When you establish business relationships, you inherit security risks from your partners. Organizations must assume that trusted partners may become compromised.
  • Traditional perimeter defenses are insufficient: Attacks that leverage legitimate business relationships and stolen credentials bypass traditional security controls that focus on external threats.
  • Credential hygiene is critical: Storing credentials in emails or documents transforms cloud platforms into credential repositories for attackers. Implement proper secrets management.
  • Detection requires behavioral analytics: Identifying this attack pattern requires sophisticated monitoring that can detect subtle anomalies in access patterns, even when technical indicators appear legitimate.
  • Incident response must consider cascading impact: When responding to cloud compromises, security teams must immediately consider whether the breach could enable attacks against partners or customers.

References

  • Google Cloud Security Blog: Workspace Security Best Practices
  • MITRE ATT&CK: Cloud Service Dashboard (T1538)
  • MITRE ATT&CK: Steal Application Access Token (T1528)
  • CISA: Cloud Security Technical Reference Architecture
  • NSA/CISA: Cloud Security Guidance

Stay updated at https://cydhaal.com — Your Daily Dose of Cyber Intelligence.
📧 Subscribe to our newsletter at https://cydhaal.com/newsletter/


Leave a Reply

Your email address will not be published. Required fields are marked *

💬 Join WhatsApp Channel 📲 Cydhaal App