llm-security
Safeguard articles tagged "llm-security" — guides, analysis, and best practices for software supply chain and application security.
120 articles
Rogue AI Agents: When Autonomous Systems Act Outside Inte...
Autonomous AI agents are gaining real access to production systems — and real incidents, from deleted databases to fabricated refunds, show what happens when they act outside intended boundaries.
Cascading Failures in Multi-Agent AI Architectures
One compromised agent can poison an entire pipeline in seconds. Heres how cascading failures spread through multi-agent AI systems, and how to contain them.
Uncontrolled Recursion in AI Agent Loops
AI agents can call themselves into runaway loops, burning thousands of dollars and crashing services. Here's why it happens and how to stop it.
Agentic Unexpected Code Execution Vulnerabilities
How AI agents with code-execution tools get hijacked by prompt injection—from the Vanna.ai RCE (CVE-2024-5565) to LangChain and MCP—and what to do about it.
Agentic AI Supply Chain Vulnerabilities
Agentic AI systems trust tools and models at runtime, not build time. Real 2024-2025 incidents show how MCP servers and AI packages become supply chain attack vectors.
Agent Goal Hijacking: Redirecting Autonomous AI Objectives
Attackers are hijacking autonomous AI agents by planting instructions in content they read—no exploit needed. Here's how it works, real 2025 incidents, and defenses.
LLM Unbounded Consumption: Resource Exhaustion Attacks
How attackers exploit token-based pricing and growing context windows to exhaust LLM compute and inflate cloud bills — and the concrete limits that stop them.
LLM Vector and Embedding Weaknesses
Embeddings aren't anonymized math — Vec2Text recovers 92% of text from vectors, and OWASP's LLM08:2025 now names inversion, poisoning, and exposed vector DBs as core AI risks.
LLM System Prompt Leakage
System prompts often hide business logic and secrets. Here's how attackers extract them, real 2023-2024 incidents, and how to stop leaks before they reach production.
Model Theft: Protecting Proprietary LLMs from Extraction ...
A $20 API attack can clone a production LLM's embeddings. Here's how model extraction works, real incidents from LLaMA to DeepSeek, and how to protect proprietary models.
Training Data Poisoning Attacks on Machine Learning Models
A $60 domain purchase or 0.001% of training tokens can silently corrupt an ML model. Here's how training data poisoning attacks work and how to defend against them.
Sensitive Information Disclosure in LLM Applications
From Samsung's ChatGPT leak to RAG pipelines with no access controls, sensitive information disclosure is now a top LLM security risk. Here's how it happens and how to stop it.
Self-healing security runs on Safeguard.
Your first fix PR is minutes away.
No sales call required, even your agent can complete the purchase over MCP.