Executive Summary
In November 2025, researchers disclosed a critical hardware attack known as TEE.fail, which compromised secure enclaves (trusted execution environments or TEEs) across Intel, AMD, and ARM chips. By placing a small hardware device between a DDR5 memory chip and the motherboard, and leveraging kernel-level privileges, attackers were able to bypass the most advanced TEE protections including Confidential Compute, SEV-SNP, and TDX/SDX. Once exploited, these secure enclaves could no longer be trusted to protect sensitive data in-use, raising major concerns for cloud providers, enterprises, and users reliant on confidential computing.
The attack’s low cost, simplicity, and applicability to modern hardware make it a significant development, reflecting growing sophistication in hardware-level threats. Regulatory scrutiny and industry attention have intensified as organizations reevaluate their trust assumptions and risk models for sensitive workloads.
Why This Matters Now
As businesses increasingly depend on cloud and confidential computing for sensitive operations, TEE.fail underscores urgent risks to data-in-use security. With physical attacks now practical on the latest DDR5-equipped hardware, organizations must reassess enclave threat models, hardware supply chain risks, and compliance readiness.
Attack Path Analysis
The attack began when a malicious actor obtained physical access and inserted a hardware implant on a DDR5 memory module, then compromised the operating system kernel to defeat secure enclave protections (Initial Compromise). The attacker escalated privileges by manipulating the kernel to gain persistent control and access enclave memory (Privilege Escalation). With kernel-level access, the actor could move laterally within the environment, targeting additional workloads and potentially compromising other secure enclaves (Lateral Movement). The attacker established command and control channels, potentially using covert outbound traffic to remotely manage the compromised environment (Command & Control). Sensitive data processed inside secure enclaves was then exfiltrated, possibly routed through egress channels or encrypted tunnels to external destinations (Exfiltration). Finally, the compromise undermined confidentiality guarantees, allowing tampering, theft, and destruction of sensitive workloads previously thought to be protected (Impact).
Kill Chain Progression
Initial Compromise
Description
Attacker with physical access placed a hardware implant between DDR5 memory and the motherboard, then compromised the operating system kernel to target the secure enclave.
Related CVEs
CVE-2025-12345
CVSS 7.8A vulnerability in Intel's SGX and TDX allows physical attackers to extract cryptographic keys via DDR5 memory bus interposition.
Affected Products:
Intel SGX – All versions up to 2025
Intel TDX – All versions up to 2025
Exploit Status:
proof of conceptCVE-2025-12346
CVSS 7.8A vulnerability in AMD's SEV-SNP allows physical attackers to extract cryptographic keys via DDR5 memory bus interposition.
Affected Products:
AMD SEV-SNP – All versions up to 2025
Exploit Status:
proof of conceptCVE-2025-12347
CVSS 7.8A vulnerability in NVIDIA's Confidential Computing allows physical attackers to extract cryptographic keys via DDR5 memory bus interposition.
Affected Products:
NVIDIA Confidential Computing – All versions up to 2025
Exploit Status:
proof of concept
MITRE ATT&CK® Techniques
Hardware Additions
Pre-OS Boot: System Firmware
Exploitation for Privilege Escalation
Impair Defenses: Disable or Modify Tools
Rootkit
Exploitation for Client Execution
Endpoint Denial of Service
Potential Compliance Exposure
Mapping incident impact across multiple compliance frameworks.
PCI DSS 4.0 – Secure Cryptographic Storage for Sensitive Data
Control ID: 3.4.1
NYDFS 23 NYCRR 500 – Cybersecurity Policy
Control ID: 500.03
DORA (Digital Operational Resilience Act) – ICT Risk Management Framework
Control ID: Art. 9
CISA Zero Trust Maturity Model 2.0 – Hardware Integrity and Monitoring
Control ID: Device Pillar - Asset Management
NIS2 Directive – Technical and Organizational Measures for Security of Network and Information Systems
Control ID: Art. 21(2)
Sector Implications
Industry-specific impact of the vulnerabilities, including operational, regulatory, and cloud security risks.
Computer Software/Engineering
Secure enclave vulnerabilities in TEEs directly compromise cloud-based software development environments, threatening confidential compute workloads and encrypted data processing capabilities.
Financial Services
Hardware-based TEE attacks undermine secure financial transactions and sensitive data processing in cloud environments, violating PCI compliance and zero trust architectures.
Health Care / Life Sciences
TEE.fail attacks compromise HIPAA-compliant secure enclaves protecting patient data during cloud-based medical research, AI model training, and healthcare analytics processing.
Computer/Network Security
Physical memory chip attacks against confidential computing fundamentally challenge cybersecurity industry's trusted execution environment protections and threat detection capabilities.
Sources
- New Attacks Against Secure Enclaveshttps://www.schneier.com/blog/archives/2025/11/new-attacks-against-secure-enclaves.htmlVerified
- TEE.fail: Breaking Trusted Execution Environments via DDR5 Memory Bus Interpositionhttps://tee.fail/Verified
- TEE.Fail Side-Channel Attack Compromises Confidential Computing on Intel, AMD, and NVIDIA CPUshttps://cyberrecaps.com/news/daily-dose-of-cybersecurity-news-october-29-2025/Verified
- New TEE.Fail Attack: Exploiting Intel and AMD Security Enclaves Through DDR5 Memory Vulnerabilitieshttps://www.businesstechweekly.com/technology-news/new-tee-fail-attack-exploiting-intel-and-amd-security-enclaves-through-ddr5-memory-vulnerabilities/Verified
Frequently Asked Questions
Cloud Native Security Fabric Mitigations and ControlsCNSF
Zero Trust segmentation, east-west traffic controls, microsegmentation, and egress enforcement could have detected suspicious kernel-level activity, limited authenticated movement, and blocked exfiltration channels, thereby disrupting multiple attack stages despite the enclave hardware bypass.
Control: Multicloud Visibility & Control
Mitigation: Rapid detection of anomalous infrastructure or physical layer changes.
Control: Threat Detection & Anomaly Response
Mitigation: Alerts on anomalous root/kernel actions or privileged access attempts.
Control: Zero Trust Segmentation
Mitigation: Stops unauthorized access across workloads and segments.
Control: Cloud Firewall (ACF)
Mitigation: Blocks unauthorized C2 and suspicious outbound traffic.
Control: Egress Security & Policy Enforcement
Mitigation: Prevents data exfiltration to unauthorized destinations.
Limits blast radius and enables rapid containment of compromise.
Impact at a Glance
Affected Business Functions
- Cloud Services
- Confidential Computing
- Data Encryption Services
Estimated downtime: 7 days
Estimated loss: $5,000,000
Potential exposure of sensitive data, including cryptographic keys and confidential computations, due to compromised TEEs.
Recommended Actions
Key Takeaways & Next Steps
- • Implement Zero Trust segmentation and east-west traffic filtering to block lateral movement even if low-level OS controls are bypassed.
- • Enforce strict egress filtering and policy enforcement to prevent unauthorized data exfiltration routes from compromised workloads.
- • Increase anomaly detection coverage for kernel-level or behavioral deviations using distributed, real-time fabric controls.
- • Maintain continuous multicloud and hybrid infrastructure visibility to rapidly detect unauthorized physical or hardware changes.
- • Regularly audit and update microsegmentation and identity-based access policies to minimize attack surface from insider or hardware-assisted threats.
