Executive Summary
Between 2021 and 2025, Amazon's threat intelligence team uncovered a multi-year cyber campaign attributed to Russia's Main Intelligence Directorate (GRU), specifically associated with APT44/Sandworm. The attackers targeted Western energy sector organizations, critical infrastructure providers, and cloud-hosted network environments by exploiting vulnerabilities and, increasingly, leveraging misconfigured network edge devices. This facilitated credential interception and lateral movement through persistent network access, with efforts focused on credential harvesting and replay against victim organizations. Amazon responded by notifying affected customers and disrupting active operations, limiting further impact.
This incident underscores the sophistication and persistence of nation-state actors in targeting vital infrastructure by adapting TTPs to minimize exposure. The campaign signals an urgent shift towards exploiting cloud and network misconfigurations rather than relying solely on zero-day vulnerabilities—a trend that broadens risk for organizations across sectors.
Why This Matters Now
This incident highlights the pressing need for improved cloud and network edge security, as attackers are increasingly targeting misconfigurations over traditional software flaws. With critical infrastructure at risk and threat actors adapting rapidly, organizations must update security controls and monitoring practices urgently to address this evolving threat landscape.
Attack Path Analysis
The GRU threat actors initially compromised cloud-hosted or on-premise edge network devices with exposed management interfaces or known vulnerabilities. After gaining device-level access, they used intercepted network traffic to harvest credentials and tokens, escalating their privileges in cloud or enterprise networks. With stolen credentials, the attackers pivoted laterally to access additional workloads and services, maintaining persistence through interactive sessions and remote management. Command and control was established using persistent outbound connections from compromised network appliances to actor-controlled IPs, blending in with legitimate traffic. Sensitive data, including credentials and potentially operational data, was exfiltrated via these covert outbound channels. The ultimate impact aimed to enable continued espionage, further credential harvesting, and strategic positioning for disruption of critical infrastructure.
Kill Chain Progression
Initial Compromise
Description
Adversaries exploited misconfigured or vulnerable edge devices hosted in the cloud (e.g., exposed management interfaces, Firebox/XTM, Confluence, Veeam), gaining foothold via remote code execution.
Related CVEs
CVE-2022-26318
CVSS 9.8An arbitrary code execution vulnerability in WatchGuard Firebox and XTM appliances allows remote attackers to execute arbitrary code.
Affected Products:
WatchGuard Firebox and XTM Appliances – < 12.5.9
Exploit Status:
exploited in the wildCVE-2021-26084
CVSS 9.8An OGNL injection vulnerability in Atlassian Confluence Server and Data Center allows remote code execution.
Affected Products:
Atlassian Confluence Server and Data Center – < 6.13.23, 6.14.0 - 7.4.11, 7.5.0 - 7.11.5, 7.12.0 - 7.12.5, 7.13.0 - 7.13.0
Exploit Status:
exploited in the wildCVE-2023-22518
CVSS 10An improper authorization vulnerability in Atlassian Confluence Data Center and Server allows unauthenticated attackers to reset Confluence and create an administrator account.
Affected Products:
Atlassian Confluence Data Center and Server – < 7.19.16, 8.0.0 - 8.3.4, 8.4.0 - 8.4.4, 8.5.0 - 8.5.3, 8.6.0
Exploit Status:
exploited in the wildCVE-2023-27532
CVSS 9.8An authentication bypass vulnerability in Veeam Backup & Replication allows unauthenticated users to access backup infrastructure hosts.
Affected Products:
Veeam Backup & Replication – < 11.0.1.1261, 12.0.0.1420
Exploit Status:
exploited in the wild
MITRE ATT&CK® Techniques
Exploit Public-Facing Application
Network Service Discovery
Adversary-in-the-Middle: LLMNR/NBT-NS Poisoning and SMB Relay
Brute Force: Credential Stuffing
Valid Accounts
Create Account
Remote Services: Remote Desktop Protocol
Network Sniffing
Potential Compliance Exposure
Mapping incident impact across multiple compliance frameworks.
PCI DSS 4.0 – Strong Authentication for Remote Access
Control ID: 8.3.1
NYDFS 23 NYCRR 500 – Cybersecurity Policy
Control ID: 500.03
DORA (Digital Operational Resilience Act) – ICT Risk Management Framework
Control ID: Article 9
CISA Zero Trust Maturity Model 2.0 – Enforce Strong Authentication and Authorization
Control ID: Identity Pillar – Authentication
NIS2 Directive – Technical and Organisational Measures
Control ID: Article 21(2)
Sector Implications
Industry-specific impact of the vulnerabilities, including operational, regulatory, and cloud security risks.
Oil/Energy/Solar/Greentech
Primary GRU target with years-long campaign exploiting network edge devices for credential harvesting, requiring enhanced encrypted traffic and zero trust segmentation capabilities.
Utilities
Critical infrastructure targeted through misconfigured edge devices and VPN concentrators, necessitating multicloud visibility and threat detection for nation-state espionage protection.
Information Technology/IT
Cloud service providers face coordinated attacks on AWS-hosted network appliances, demanding inline IPS and egress security to prevent credential replay operations.
Telecommunications
Telecom providers specifically targeted in credential replay operations across multiple regions, requiring east-west traffic security and secure hybrid connectivity measures.
Sources
- Amazon Exposes Years-Long GRU Cyber Campaign Targeting Energy and Cloud Infrastructurehttps://thehackernews.com/2025/12/amazon-exposes-years-long-gru-cyber.htmlVerified
- Amazon Threat Intelligence identifies Russian cyber threat group targeting Western critical infrastructurehttps://aws.amazon.com/blogs/security/amazon-threat-intelligence-identifies-russian-cyber-threat-group-targeting-western-critical-infrastructure/Verified
- CVE-2023-22518 - Improper Authorization Vulnerability In Confluence Data Center and Serverhttps://confluence.atlassian.com/security/cve-2023-22518-improper-authorization-vulnerability-in-confluence-data-center-and-confluence-server-1311473907.htmlVerified
- Russian GRU Cyber Actors Targeting Western Logistics Entities and Tech Companieshttps://www.cisa.gov/news-events/alerts/2025/05/21/russian-gru-cyber-actors-targeting-western-logistics-entities-and-tech-companiesVerified
Frequently Asked Questions
Cloud Native Security Fabric Mitigations and ControlsCNSF
Implementing CNSF zero trust segmentation, east-west traffic controls, encryption-in-transit, and robust egress policy enforcement would have significantly constrained this attack at multiple stages, blocking lateral movement, credential harvesting, and data exfiltration. CNSF’s visibility and inline policy enforcement are critical to detecting early-stage compromise and halting malicious persistence in cloud-hosted edge environments.
Control: Cloud Firewall (ACF)
Mitigation: Blocked unauthorized external access to management interfaces.
Control: Encrypted Traffic (HPE)
Mitigation: Prevented credential sniffing and interception.
Control: Zero Trust Segmentation
Mitigation: Restricted intra-cloud movement to only explicitly allowed identities and services.
Control: Egress Security & Policy Enforcement
Mitigation: Blocked or alerted on unauthorized outbound C2 connections.
Control: Inline IPS (Suricata)
Mitigation: Detected and blocked known data exfiltration signatures.
Real-time alerting on anomalous behavior enabled rapid incident response.
Impact at a Glance
Affected Business Functions
- Energy Production
- Cloud Services
- Network Management
Estimated downtime: 7 days
Estimated loss: $5,000,000
Potential exposure of sensitive operational data and customer information due to unauthorized access and credential harvesting.
Recommended Actions
Key Takeaways & Next Steps
- • Audit and restrict exposure of all edge device management interfaces and enforce cloud firewall policies.
- • Deploy mandatory encrypted-in-transit controls (e.g., MACsec/IPsec) for all sensitive cloud and network traffic.
- • Apply granular zero trust segmentation and least-privilege policies for intra-cloud and hybrid network flows.
- • Enforce centralized egress filtering and inline IPS to prevent covert C2 and exfiltration attempts.
- • Monitor for credential harvesting patterns and anomalous activity with continuous threat detection and rapid response automation.



