AI & LLM Security Test
Tests chatbots, RAG pipelines and AI agents against the OWASP Top 10 for LLM Applications: prompt injection, sensitive data disclosure, insecure tool use and excessive agency.
Indicative starting price $4,950 · final price confirmed in your quote
What we test
- Direct and indirect prompt injection
- Training and RAG data leakage
- Tool and agent permission abuse
- Output handling and jailbreaks
What you get
- Model-behaviour findings
- Guardrail recommendations
- Executive summary
- Free retest
How the engagement runs
Scope and written authorisation first, then reconnaissance, testing, live escalation of critical issues, reporting and a free retest. Read the full methodology.