OBSERRA TRUST CENTER
Responsible AI Policy
AI use must remain governed, explainable, proportionate to risk, and subject to appropriate human oversight.
Governance Principles
- AI use is evaluated for business purpose, data sensitivity, legal obligations, security risk, and potential harm.
- Human review is required for consequential recommendations, approvals, or actions where appropriate.
- AI generated output is treated as decision support and must be validated before material use.
Data and Security
- Sensitive client information is not submitted to unapproved AI services.
- Access, logging, retention, and model usage controls are applied according to risk and contractual requirements.
- Prompt injection, data leakage, model abuse, and unauthorized automation risks are considered in design and operation.
Accountability
- AI recommendations should be traceable to relevant evidence, assumptions, and confidence where practical.
- Users remain accountable for final decisions and authorized actions.
- Suspected AI related security or governance issues should be reported to info@obserrallc.com.