owasp-llm-top-10
Safeguard articles tagged "owasp-llm-top-10" — guides, analysis, and best practices for software supply chain and application security.
22 articles
The emerging role of the AI security engineer
OWASP's 2025 LLM Top 10 ranks prompt injection #1 and calls it structurally unfixable by parameterization — a signal that AppSec skills alone no longer cover the job.
The guardrail gap in low-code agentic AI platforms
Low-code AI builders let business users wire agents to live connectors in minutes — but most ship without per-tool scoping, approval gates, or audit trails.
Prompt Injection vs Traditional Injection Attacks: A Tech...
SQL injection was solved by separating code from data. Prompt injection can't be, because in an LLM they share one channel. Here's the technical comparison, with real exploits and dates.
Distinguishing Model Risk from Application Risk in Agenti...
Model flaws and application flaws in AI agents cause different breaches and need different fixes. Real incidents show where each risk actually lives — and how to test for both.
What is LLM Security
LLM security protects model weights, prompts, outputs, and the AI supply chain from injection, leakage, and compromise—here's what it covers and how to defend it.
What is Prompt Injection
Prompt injection is OWASP's #1 LLM risk. Learn how it works, real CVEs like EchoLeak, and how to detect and defend against it.
Direct vs Indirect Prompt Injection
Direct and indirect prompt injection are different attack vectors with different blast radii. Real 2025 CVEs like EchoLeak show why the distinction matters.
What is Training Data Poisoning
Training data poisoning corrupts an ML model's training data to plant hidden backdoors. Learn how it works, real incidents, and how to detect it.
What is Insecure Output Handling in LLMs
Insecure output handling lets LLM-generated text execute code, alter queries, or render unsanitized HTML — a real, exploitable OWASP LLM05:2025 risk.
What is Sensitive Information Disclosure in LLMs
LLM sensitive information disclosure leaks training data, prompts, and secrets through model outputs. Real incidents, causes, and defenses explained.
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.