1. Scope
This policy applies to AI-assisted features, agents, automations, prompts, generated content and actions performed through myMulti or connected services.
2. AI transparency
myMulti is designed to make clear when a person is interacting directly with an AI system where required by applicable law. Customers deploying myMulti to interact with their own customers must preserve required AI disclosures and may not intentionally misrepresent automated interactions as human where such disclosure is legally required.
3. Human oversight and authorization
AI systems can make mistakes. Customers must apply human review and appropriate authorization thresholds before relying on AI for material or irreversible actions, especially financial transfers, legal commitments, employment decisions, health or safety decisions, regulated advice, identity decisions or other high-impact activities.
4. Prohibited uses
- Illegal activity, fraud, deception, impersonation or evasion of legal obligations.
- Unauthorized access, credential theft, malicious code or disruption of systems.
- Using personal data without a valid legal basis or required notice/consent.
- Discriminatory or unlawful high-impact decision-making.
- Generating or distributing content that infringes rights or violates applicable law.
- Using myMulti to make commitments or transactions beyond the authority granted by the relevant business.
5. Data minimization
Users should provide only data necessary for the requested task and avoid unnecessary sensitive information. Customers remain responsible for determining whether their use case requires consent, contractual safeguards, a DPIA or other assessment.
6. Model and tool providers
myMulti may route tasks through different AI models or software tools. Model choice can affect capabilities, latency, data location and limitations. Applicable providers are disclosed in the Subprocessor Register when they process personal data on behalf of myMulti or its customers.
7. Actions and confirmations
Where technically supported, customers should use confirmation, permissioning, spending limits, role-based access and audit logs for higher-risk actions. myMulti may block or require additional verification for suspicious or sensitive requests.
8. Enforcement
myMulti may restrict, suspend or terminate use that materially violates this policy or creates a security, legal or safety risk. Questions may be sent to legal@mymulti.ai.