Service
Crisis and vulnerable-user response testing
People tell chatbots things they would not tell a company. We test how your assistant responds when a user is distressed, in danger or under 18, including when the signal is buried in an ordinary request.
What we test
- Self-harm and crisis signals, stated directly and indirectly
- Users who say or imply they are children
- Dependence, flattery and pressure over long conversations
- Hand-off to a person and to appropriate help
What you get
Results graded against a written protocol, the conversations that went wrong, and a protocol you can publish if you do not have one.
Where the fix is in the setup (a prompt, a permission, a retrieval or guardrail setting) we make it and run the failed cases again. The test suite is yours to keep.
Who asks for this
Consumer apps, companion and wellbeing products, education tools, and any support bot with a large public audience.
Contact
Tell us what your assistant does. We'll tell you how we'd test it.
Other work
- Prompt injection testing for AI agents
- Third-party LLM red teaming
- Independent AI assistant assessment report
- Chatbot hallucination and accuracy audit
- EU AI Act transparency testing for chatbots
- Custom evaluation datasets and grading rubrics
- AI agent tool and permissions review
- System prompt and data leakage testing
- AI vendor due diligence testing
- Regression testing for model upgrades