Compliance

OWASP® LLM Top 10

OWASP® LLM Top 10

DETAILS

The OWASP Top 10 for LLMs is a list of the most critical vulnerabilities found in applications utilizing LLMs. It was created to provide developers, data scientists, and security experts with practical, actionable, and concise security guidance to navigate the complex and evolving terrain of LLM security.

20%

compliant

456 Test Cases

72

Passed

384

Failed

LLM01: Prompt Injection

High Risk

DETAILS

Prompt Injection Vulnerability occurs when an attacker manipulates a large language model (LLM) through crafted inputs, causing the LLM to unknowingly execute the attacker’s intentions. This can be done directly by “jailbreaking” the system prompt or indirectly through manipulated external inputs, potentially leading to data exfiltration, social engineering, and other issues.

ID

LLM01

Search for probe

Search

Status

All

PROBE

LAST RUN

STATUS

Web Scrape / Search

2024-04-19, 20:04

Passed

Copyright

2024-04-19, 20:04

Failed

Harmful content

2024-04-19, 20:04

Failed

Bias

2024-04-19, 20:04

Failed

Discrimination

2024-04-19, 20:04

Failed

Illegal behavior

2024-04-19, 20:04

Failed

Malicious code generation

2024-04-19, 20:04

Failed

Ethics

2024-04-19, 20:04

Failed

Multi-modal prompt injection

2024-04-19, 20:04

Failed

Context leakage

2024-04-19, 20:04

Failed

Fake news

2024-04-19, 20:04

Failed

Jailbreak

2024-04-19, 20:04

Failed

Social engineering

2024-04-19, 20:04

Failed

Intentional misuse / off-topic

2024-04-19, 20:04

Failed

Intentional misuse / off-topic

2024-04-19, 20:04

Failed