Eresus research, advisory, and security news
We collect writing, advisories, and current-event analysis around AI security, the MCP ecosystem, application security, and real attack chaining here.
Latest Posts
Hacking Humans: Social Engineering and the Psychology
Social engineering engagements are the most exciting and heart pumping. It doesn’t begin at the badge reader or the front desk. The access occurs when someone makes a decision.
Transitive Model Threat Detected with Unsafe Model Dependency
Highlighting critical interconnected runtime threats targeting definitively unsafe Artificial Intelligence components distinctly imported during ML...
Transitive Model Threat Detected with A Suspicious Model Dependency
Discover why transitive model threats and suspicious artificial intelligence dependencies put your AI supply chain at risk. Learn how Eresus Sentinel...
TensorFlow SavedModel: Suspicious Operator Execution at Runtime
What a PAIT-TF-302 finding means in a TensorFlow SavedModel, how to separate suspicious runtime behavior from a confirmed exploit, and how to quarantine and retest the artifact.
The Simplest Bug is the Deadliest: Remote Code Execution (RCE) via Pickle in Machine Learning
Sometimes the simplest bugs are the most dangerous—especially when they’ve been hiding in plain sight. In the world of Machine Learning (ML), data ...
Perseus Android Banking Malware: Accessibility Abuse and Notes-App Theft
What the Perseus Android banking trojan does with Accessibility Services, why note-taking apps are a target, and how users and enterprises can contain the risk.
TensorFlow SavedModel Contains Unsafe Operator Execution at Model Run Time
Critical execution vulnerability identifying specifically unsafe logical operators strictly executing natively during standard TensorFlow prediction...
TensorFlow Custom Operator Injection (Graph Execution)
Identifies backdoor threats leveraging embedded malformed computational nodes (Custom Operators) within TensorFlow models to silently trigger execution...
TorchScript Model Arbitrary Code Execution Suspected at Model Load Time
Highlights suspicious computational graph behavior evaluating directly indicative of load-time remote code execution attempts within AI infrastructure....