Attackers Leverage Pandoc SSRF Vulnerability CVE-2025-51591 to Breach AWS IMDS
Executive Summary
In September 2025, threat actors exploited a newly disclosed Server-Side Request Forgery (SSRF) vulnerability in the open-source Linux utility Pandoc (CVE-2025-51591), targeting Amazon Web Services (AWS) cloud environments. The attackers leveraged the flaw to send unauthorized requests to the AWS Instance Metadata Service (IMDS), allowing them to obtain EC2 role credentials and elevate cloud permissions. Security researchers, including Wiz, observed active exploitation in the wild, leading to unauthorized access and potential data exfiltration from affected AWS infrastructure. Organizations relying on Pandoc as part of their cloud automation workflows face heightened risk of credential compromise and lateral movement across accounts.
This incident underscores a fast-evolving cloud threat landscape, where attackers exploit supply-chain and open-source vulnerabilities to traverse trusted infrastructure and target sensitive identity and metadata services. The rapid weaponization of CVE-2025-51591 mirrors the broader trend of SSRF attacks on cloud metadata, driving urgent calls for proactive detection, segmentation, and credential management in multi-cloud environments.
Why This Matters Now
Cloud supply chain and open-source component vulnerabilities are now being rapidly weaponized by attackers to breach cloud-native environments, access sensitive IMDS endpoints, and steal dynamic IAM credentials. As organizations expand automation and rely on tools like Pandoc, the window between vulnerability disclosure and exploitation is shrinking—making swift detection, zero trust segmentation, and robust egress policies urgently necessary to block credential theft and prevent widespread compromise.
Attack Path Analysis
Attackers exploited the Pandoc SSRF vulnerability (CVE-2025-51591) to gain initial access to an AWS EC2 environment. Leveraging SSRF, they escalated privileges by accessing sensitive AWS Instance Metadata Service (IMDS) endpoints to harvest IAM credentials. With valid credentials, the adversary potentially moved laterally within the cloud environment, seeking additional assets. The attackers established command and control, maintaining access and preparing for further operations via outbound channels. Stolen credentials were exfiltrated through outbound network connections. Ultimately, the attackers obtained the ability to impact cloud resources, possibly enabling further abuse or destructive actions.
Kill Chain Progression
Initial Compromise
Description
Exploited a Pandoc SSRF vulnerability to gain access to a public-facing workload in the cloud environment.
Related CVEs
CVE-2025-51591
CVSS 6.5A Server-Side Request Forgery (SSRF) vulnerability in JGM Pandoc v3.6.4 allows attackers to inject crafted iframes, potentially compromising infrastructure by accessing internal network resources.
Affected Products:
JGM Pandoc – 3.6.4
Exploit Status:
exploited in the wild
MITRE ATT&CK® Techniques
Exploit Public-Facing Application
Data from Information Repositories: Cloud Storage Object
Unsecured Credentials: Credentials in Files
Valid Accounts: Cloud Accounts
Cloud Service Discovery
Exfiltration Over C2 Channel
Browser Session Hijacking
Potential Compliance Exposure
Mapping incident impact across multiple compliance frameworks.
PCI DSS v4.0 – Log and Monitor All Access to System Components and Cardholder Data
Control ID: 10.2.5
NYDFS 23 NYCRR 500 – Cybersecurity Policy
Control ID: 500.03
DORA (EU Digital Operational Resilience Act) – ICT Risk Management Framework
Control ID: Article 9
CISA Zero Trust Maturity Model (ZTMM 2.0) – Protect All Credentials and Enforce Strong Authentication
Control ID: Identity Pillar: Credentials Protection
NIS2 Directive – Incident Handling and Vulnerability Management
Control ID: Article 21.2(d)
Sector Implications
Industry-specific impact of the vulnerabilities, including operational, regulatory, and cloud security risks.
Information Technology/IT
Critical exposure through widespread Pandoc usage in cloud infrastructure, enabling SSRF attacks against AWS IMDS to compromise IAM credentials and cloud security.
Financial Services
High-risk sector utilizing AWS cloud services extensively, vulnerable to credential theft via CVE-2025-51591 exploitation affecting compliance requirements and sensitive data protection.
Health Care / Life Sciences
Significant threat to HIPAA-compliant cloud deployments using Pandoc, with SSRF attacks potentially exposing patient data through compromised EC2 IAM credentials.
Government Administration
Government cloud infrastructure at risk from Pandoc SSRF vulnerability, enabling unauthorized access to sensitive systems through AWS metadata service exploitation.
Sources
- Hackers Exploit Pandoc CVE-2025-51591 to Target AWS IMDS and Steal EC2 IAM Credentialshttps://thehackernews.com/2025/09/hackers-exploit-pandoc-cve-2025-51591.htmlVerified
- CVE-2025-51591 Impact, Exploitability, and Mitigation Steps | Wizhttps://www.wiz.io/vulnerability-database/cve/cve-2025-51591Verified
- CVE-2025-51591 : A Server-Side Request Forgery (SSRF) in JGM Pandoc v3.6.4 allows attackers to gahttps://www.cvedetails.com/cve/CVE-2025-51591/Verified
Frequently Asked Questions
Cloud Native Security Fabric Mitigations and ControlsCNSF
Network segmentation, east-west traffic controls, egress policy enforcement, and threat detection provided by CNSF and zero trust frameworks would have significantly limited the attacker's ability to exploit, move laterally, or exfiltrate data at multiple points in the kill chain.
Control: Inline IPS (Suricata)
Mitigation: Attempted SSRF exploit detected and blocked inline.
Control: Zero Trust Segmentation
Mitigation: Access to metadata endpoints restricted by microsegmentation.
Control: East-West Traffic Security
Mitigation: Lateral movement between services detected and blocked.
Control: Threat Detection & Anomaly Response
Mitigation: Anomalous or unauthorized outbound communication detected and alerted.
Control: Egress Security & Policy Enforcement
Mitigation: Unauthorized data exfiltration attempt blocked at the network edge.
Unusual resource or identity actions detected across cloud environments.
Impact at a Glance
Affected Business Functions
- Cloud Infrastructure Management
- Data Security
Estimated downtime: 3 days
Estimated loss: $50,000
Potential unauthorized access to AWS IAM credentials, leading to possible data breaches and unauthorized access to cloud resources.
Recommended Actions
Key Takeaways & Next Steps
- • Enforce strict zero trust segmentation to prevent unauthorized workload-to-workload and metadata endpoint access.
- • Deploy inline intrusion prevention to block exploitation attempts of known vulnerabilities like Pandoc SSRF (CVE-2025-51591).
- • Implement robust egress control policies to restrict outbound traffic and prevent credential exfiltration.
- • Continuously monitor east-west and egress flows for signs of lateral movement, C2, and data leak attempts.
- • Centralize multicloud visibility for unified policy enforcement, alerting, and rapid incident response across all environments.
