Test an AI assistant for information disclosure
Assess whether a customer-facing chatbot can be induced to disclose its system prompt, embedded secrets or other restricted information.
Skills covered
- 01Testing whether an LLM discloses its system prompt
- 02Recognising sensitive information that must not be included in prompts
- 03Reducing information-disclosure risk in AI assistants
More simulations
10- 01Identify a phishing emailRed team · Phishing
- 02Identify a fraudulent support callSocial engineering · Voice
- 03Detect indirect prompt injectionAI red-teaming · LLM
- 04Verify a suspected deepfake requestSocial engineering · Synthetic media
- 05Identify invoice-redirection fraudFraud · BEC
- 07Recognise social-engineering techniquesSocial engineering · Human factors
- 08Assess an unapproved cloud applicationAttack surface · Shadow IT
- 09Assess what a social-media post revealsOSINT · Oversharing
- 10Identify credential-stuffing riskCredentials · Account security
- 11Manage the first hour of an incidentIncident response