BlogAI Security

OWASP Top 10 for LLM Applications: A Developer's Checklist

LLM applications introduce entirely new categories of security risk — prompt injection, data leakage through model outputs, excessive tool permissions, and insecure plugin designs. The OWASP Top 10 for LLM Applications provides a framework for understanding and mitigating these risks. Here is the practical developer checklist.

22 min readUpdated June 2026

Why LLM Security Matters

LLM applications are different from traditional software. They accept natural language input that can be crafted to manipulate behavior, they generate outputs that may contain sensitive data or malicious content, and they increasingly have agency — the ability to execute tools, call APIs, and modify data. Each of these capabilities creates attack surface that did not exist before.

90%+

of LLM applications are vulnerable to at least one form of prompt injection

5x

increase in AI-related security incidents reported between 2023–2025

40%

of organizations deploying AI agents have no security testing for them

The OWASP Top 10 for LLM Applications

LLM01

Prompt Injection

Critical

Attackers craft inputs that override system instructions, causing the LLM to ignore its original purpose. Includes direct injection (user input) and indirect injection (injected via external data sources).

Input sanitization, prompt hardening, output validation, privilege separation between system and user prompts
LLM02

Insecure Output Handling

High

LLM outputs are passed to downstream systems without validation, enabling XSS, SSRF, SQL injection, or command injection through model-generated content.

Treat LLM output as untrusted input. Apply output encoding, sanitization, and validation before rendering or executing.
LLM03

Training Data Poisoning

High

Malicious data in training or fine-tuning datasets introduces backdoors, biases, or vulnerabilities into the model's behavior.

Data provenance tracking, input validation for training data, anomaly detection in model outputs, regular model evaluation
LLM04

Model Denial of Service

Medium

Crafted inputs that cause excessive resource consumption — extremely long prompts, recursive tool calls, or inputs that trigger expensive computation.

Input length limits, rate limiting, timeout enforcement, cost monitoring, request quotas per user
LLM05

Supply Chain Vulnerabilities

High

Compromised model weights, poisoned training pipelines, vulnerable dependencies in the LLM application stack, or malicious plugins.

Model provenance verification, dependency scanning, plugin sandboxing, SBOM for AI components
LLM06

Sensitive Information Disclosure

High

LLM reveals sensitive data from training data, RAG context, or system prompts. Includes PII leakage, credential exposure, and system prompt extraction.

PII detection in outputs, data classification, output filtering, system prompt protection, RAG access controls
LLM07

Insecure Plugin Design

High

Plugins/tools with excessive permissions, missing input validation, or no authentication. An attacker who controls LLM output can abuse poorly designed plugins.

Least-privilege plugin permissions, input validation on all plugin parameters, human-in-the-loop for destructive actions
LLM08

Excessive Agency

Critical

LLM agents with too many permissions — ability to execute code, modify databases, send emails, or access external systems without appropriate guardrails.

Principle of least privilege for all tools, confirmation prompts for high-impact actions, scope limitations, action logging
LLM09

Overreliance

Medium

Users or systems trusting LLM outputs without verification, leading to incorrect code, wrong facts, or harmful recommendations being acted upon.

Output validation, confidence scoring, human review for critical decisions, clear AI disclosure
LLM10

Model Theft

Medium

Unauthorized access to proprietary model weights, fine-tuning data, or system prompts through API abuse, side-channel attacks, or insider threats.

Access controls on model endpoints, rate limiting, watermarking, monitoring for extraction patterns

Automated LLM Security Testing

Manual testing for LLM vulnerabilities does not scale. Automated tools can test for prompt injection, PII leakage, harmful content generation, and excessive agency continuously as part of your CI/CD pipeline.

Agentic Radar

Scans AI agent workflows (LangGraph, CrewAI, OpenAI Agents) for security issues including prompt injection vectors, tool permission misconfigurations, and data flow vulnerabilities.

Prompt Injection Detection

Automated testing with adversarial prompts — jailbreak attempts, indirect injection payloads, and system prompt extraction techniques.

PII Leakage Scanning

Test LLM outputs for personally identifiable information, credit card numbers, API keys, and other sensitive data that should not appear in responses.

Prompt Hardening

Automated generation of hardened system prompts that are more resistant to injection attacks, with defensive instructions and output constraints.

Secure Your AI Applications with TigerGate

TigerGate's AI Scanner detects prompt injection vectors, PII leakage, and agent misconfigurations across LangGraph, CrewAI, OpenAI Agents, and more — mapped to the OWASP Top 10 for LLMs.