127 indexed

COMPLIANCECompliance controls

127 controls across 14 compliance frameworks, grouped by framework. For cross-framework Jaccard overlap see /explore/crosswalk. Authored by Adam Lundqvist.

10 in OWASP LLM TOP10 · 127 total

IDTitleSummary
OWASP_LLM_TOP10-LLM01Prompt Injection
OWASP LLM TOP10pentest:high
Prompt injection vulnerabilities occur when user prompts alter the LLM's behaviour or output in unintended ways. These inputs can affect the model even if they…
OWASP_LLM_TOP10-LLM02Sensitive Information Disclosure
OWASP LLM TOP10pentest:high
LLMs can expose sensitive information through their outputs, including personally identifiable information (PII), proprietary algorithms, confidential business…
OWASP_LLM_TOP10-LLM03Supply Chain
OWASP LLM TOP10pentest:high
LLM supply chains are vulnerable to integrity failures, particularly in training data, models, and deployment platforms. Risks include compromised pretrained m…
OWASP_LLM_TOP10-LLM04Data and Model Poisoning
OWASP LLM TOP10pentest:high
Data and model poisoning attacks occur when an attacker manipulates the pretraining, fine-tuning, or embedding data of the LLM to introduce vulnerabilities, ba…
OWASP_LLM_TOP10-LLM05Improper Output Handling
OWASP LLM TOP10pentest:high
Improper output handling occurs when LLM-generated output is passed downstream to other systems without validation, sanitisation, or context-aware escaping. Th…
OWASP_LLM_TOP10-LLM06Excessive Agency
OWASP LLM TOP10pentest:high
Excessive agency arises when LLM-based systems are granted excessive functionality, permissions, or autonomy. Damaging actions can occur in response to unexpec…
OWASP_LLM_TOP10-LLM07System Prompt Leakage
OWASP LLM TOP10pentest:high
System prompts contain configuration, instructions, and sometimes sensitive data (credentials, internal endpoints, business logic) that should not be exposed. …
OWASP_LLM_TOP10-LLM08Vector and Embedding Weaknesses
OWASP LLM TOP10pentest:high
Vulnerabilities in vector databases and embedding stores used by RAG applications. Includes unauthorised access to embedding stores, cross-tenant data leakage …
OWASP_LLM_TOP10-LLM09Misinformation
OWASP LLM TOP10pentest:medium
LLMs can generate plausible-but-false outputs (hallucinations) that downstream consumers rely upon. The risk increases when the LLM is used for high-stakes dec…
OWASP_LLM_TOP10-LLM10Unbounded Consumption
OWASP LLM TOP10pentest:high
Unbounded consumption refers to model invocations that consume excessive resources — compute, memory, tokens, or external API quota — leading to denial of serv…
Sourced from EUR-Lex (DORA, NIS2, GDPR, AI Act, CRA), ISO, NIST, OWASP, CIS, PCI SSC, ENISA TIBER-EU. Curated by Adam Lundqvist, Founder at SQUR.
Compliance controls — by framework | SQUR Knowledge Base