data-poisoning
Safeguard articles tagged "data-poisoning" — guides, analysis, and best practices for software supply chain and application security.
5 articles
AI Supply Chain Security: Securing Models and Datasets
Your AI supply chain is not just your npm dependencies anymore. It is the models you download, the weights you load, and the datasets you train on — and each is an attack surface most software security programs have never inventoried.
AI Data Poisoning Defense: Protecting Models from Tainted Data
You do not need to corrupt most of a training set to backdoor a model — recent research suggests a small, near-constant number of poisoned documents can be enough. Defense starts with treating data as a dependency.
Training Data Poisoning: Pipeline Defenses
A senior engineer's guide to training data poisoning defenses in 2026, from split-learning detection to provenance attestation and continuous pipeline monitoring.
Securing the fine-tuning pipeline against injected malici...
Fine-tuning pipeline security is now an AI supply chain priority: as few as 250 poisoned documents can backdoor a model, and LoRA adapters make it easy to hide.
How data poisoning attacks corrupt LLM behavior during tr...
A single expired domain and $60 can poison a training set. Here's how data poisoning attacks corrupt LLM behavior — and how Safeguard verifies training data before it ships.