3.7OWASP LLM Top 10 Summary Table

#	Vulnerability	Severity	Testability	Primary Testing Approach
LLM01	Prompt Injection	Critical	High	Red team payloads, fuzzing
LLM02	Insecure Output Handling	High	High	Output sanitization tests
LLM03	Training Data Poisoning	High	Low	Bias probing, output analysis
LLM04	Model Denial of Service	Medium	High	Resource exhaustion tests
LLM05	Supply Chain	High	Medium	Dependency scanning, provenance
LLM06	Sensitive Info Disclosure	Critical	High	PII detection, data probing
LLM07	Insecure Plugin Design	High	High	Permission boundary tests
LLM08	Excessive Agency	High	Medium	Action chain analysis
LLM09	Overreliance	Medium	Medium	Factual accuracy benchmarks
LLM10	Model Theft	Medium	Low	Rate limiting, anomaly detection

Every entry in this table should have corresponding automated tests in your CI pipeline. The OWASP LLM Top 10 is the minimum security test coverage for any AI application.