Skip to main content
Xoxoday implements controls to prevent sensitive data disclosure through AI features: (1) Input filtering: Prompts sent to frontier model APIs are constructed using predefined templates with controlled data inclusion.
Xoxoday implements controls to prevent sensitive data disclosure through AI features: (1) Input filtering: Prompts sent to frontier model APIs are constructed using predefined templates with controlled data inclusion. Sensitive personal data (e.g., salary details, identification numbers) is excluded from API calls by design, (2) Output filtering: AI-generated responses are reviewed/filtered before being displayed to users, (3) Data minimization: Only the minimum data necessary for the AI feature’s function is included in API calls, (4) Access scoping: The AI features operate within the authenticated user’s data access scope and cannot surface data the user is not authorized to see, (5) Provider agreements: Xoxoday’s agreements with AI providers prohibit data retention and cross-customer data exposure. Learn more: Empuls Help Centre

Is the AI / LLM model tested to prevent prompt injections?

Xoxoday applies input validation, filtering, and sanitization controls to mitigate prompt injection risks…

Is the AI / LLM model tested to prevent model poisoning?

Not directly applicable (Xoxoday does not train or host AI models). Xoxoday does not train, retrain, or host AI/LLM models…