AI Infrastructure
AI infrastructure explores GPUs, compute, training, inference and large-scale deployment. Articles follow platform evolution, architectural bottlenecks and the security implications of AI workloads.
Backdoored LiteLLM on PyPI: Malware Triggers on Python Startup Alone
On March 24, 2026, two malicious LiteLLM versions exfiltrated credentials from over 50 categories via a .pth mechanism. The compromise…
Intel and AMD Patch 70 Flaws: Two Critical CVSS 9.3 and 9.2 Bugs in GPU Drivers
On May 13, 2026, Intel and AMD released 28 advisories covering 69 vulnerabilities. Two critical flaws hit chip software drivers, not t…
RufRoot: The AI Vulnerability That Survives the Patch — 233 Tools Exposed and Persistent Memory Poisoning
CVE-2026-59726 in Ruflo exposes 233 MCP tools without authentication, enabling RCE, LLM API key theft, and persistent memory poisoning…
NVIDIA NVTabular: RCE via Pickle, CVE-2026-24237 Rated CVSS 7.8
A deserialization flaw in NVIDIA NVTabular enables remote code execution through malicious pickle files. User interaction is required;…
EncForge: JadePuffer Hits Irrecoverable AI Models With Agentic Ransomware
The agentic threat actor JadePuffer has deployed EncForge, ransomware purpose-built for AI/ML assets. Encrypted models cannot be recov…
In Internal Test, OpenAI AI Agent Breaches Hugging Face to Obtain ExploitGym Solutions
During a controlled offensive cyber evaluation, OpenAI models with reduced cyber refusals escaped a sandbox and compromised Hugging Fa…
Langflow: CISA Orders 72-Hour Patch for Pre-Auth RCE as Root
CVE-2026-0770 enables unauthenticated remote code execution as root in Langflow. CISA mandates remediation by July 24, 2026 for federa…
NVIDIAScape: Container Escape in Three Lines of Code in the NVIDIA Toolkit
CVE-2025-23266, rated CVSS 9.0, affects 37% of AI cloud environments. An old-school bug in the NVIDIA Container Toolkit enables privil…
TrapDoor: 34+ Malicious Packages Turn AI Assistants Into Insider Threats
The TrapDoor campaign has distributed over 34 packages across npm, PyPI, and Crates.io with multi-stage payloads and hidden instructio…
NVIDIA NeMo Framework: RCE Vulnerability in ML Checkpoints
An unsafe deserialization flaw in NVIDIA NeMo Framework checkpoints enables remote code execution. User interaction is required, but t…
Ollama Zero-Day DoS: downloadBlob Bug Puts Local AI Servers at Risk
ZDI has disclosed a zero-day vulnerability in Ollama enabling unauthenticated remote denial-of-service attacks via the downloadBlob fu…
CISA Orders 3-Day Patch for CVE-2026-55255 in Langflow
An IDOR in Langflow's /api/v1/responses endpoint lets authenticated attackers steal LLM and cloud credentials from other users' flows.…