machine-learning
Safeguard articles tagged "machine-learning" — guides, analysis, and best practices for software supply chain and application security.
10 articles
PyTorch Security Guide (2026)
PyTorch is the dominant deep-learning framework for research and production — and its torch.load remote-code-execution history makes loading a model checkpoint one of the most security-sensitive operations in modern ML.
TensorFlow Security Guide (2026)
TensorFlow is one of the most widely deployed machine-learning frameworks — and its history of model-deserialization RCE and crafted-tensor memory bugs makes its version and loading habits genuinely security-relevant.
AI Models in Your Supply Chain: The Security Risks Nobody Talks About
AI/ML models are the new open source libraries. Here's why your supply chain security strategy needs to account for model provenance, poisoning, and compliance.
Vulnerability Scanning for AI Models: A New Frontier
AI models ship with dependencies, use vulnerable libraries, and introduce novel attack surfaces. Traditional scanning is not enough.
Inside DeepCode AI: how Snyk Code's ML models are trained...
How Snyk's DeepCode AI turns millions of open-source commit fixes into the symbolic-AI and ML models powering Snyk Code's vulnerability detection and autofixes.
AI SBOMs and Model Cards: Building Transparency Into the AI Supply Chain
As AI models become critical software components, the need for AI-specific SBOMs and model cards grows urgent. How the industry is extending supply chain transparency to machine learning pipelines.
Securing ML Model Serving Infrastructure
Model serving infrastructure is a growing attack surface that most security teams overlook. From model poisoning to inference API abuse, here are the risks and how to address them.
Open Source AI Model Security: The Emerging Threat Landscape
As open source AI models proliferate, their security implications extend far beyond traditional software vulnerabilities. Model poisoning, supply chain tampering, and unsafe deserialization create new attack surfaces.
AI Model Poisoning: Detection Techniques for the Software Supply Chain
Poisoned AI models are a supply chain threat that traditional security tools can't detect. Here are the emerging techniques for identifying compromised models.
SBOMs for AI/ML Models: Why Machine Learning Needs a Bill of Materials
As AI models become critical infrastructure, the need for transparency about their components, training data, and dependencies grows urgent. Emerging standards are beginning to address this gap.