Purpose
This policy explains important limitations and responsibilities associated with AI functionality provided through NaviCS AI.
AI Output
AI-generated output may contain:
- factual errors;
- incomplete information;
- inappropriate recommendations;
- biased or unexpected results;
- outdated information;
- incorrect interpretations.
AI output should therefore be reviewed before being used for consequential purposes.
Human Oversight
Customers are responsible for establishing appropriate human oversight for AI-powered workflows.
Where an action may materially affect an individual, organization, financial position, legal rights, employment, safety, or access to services, customers should establish appropriate review and escalation procedures.
Instructions and Configuration
Digital Employees and AI agents operate according to configured instructions, prompts, tools, workflows, knowledge sources, permissions, and integrations.
Customers are responsible for ensuring that these configurations are appropriate for their intended use.
AI Providers
The Services may use third-party AI providers.
AI-provider availability, capabilities, model behavior, processing terms, and geographic availability may vary.
Under legal review. Final provider-specific processing and training language.
Confidential Information
Customers should avoid providing confidential or sensitive information to an AI workflow unless the applicable configuration and contractual terms are appropriate for that information.
Credentials and authentication secrets must be managed through designated credential-management mechanisms rather than ordinary prompts or application content.
Accuracy and Verification
NaviCS AI does not guarantee that AI-generated output is accurate or complete.
Customers should independently verify material information before relying on AI output.
Prohibited AI Uses
AI functionality must not be used to:
- facilitate unlawful activity;
- generate malicious code for unauthorized attacks;
- impersonate individuals for fraud;
- circumvent security controls;
- unlawfully manipulate or deceive individuals;
- make prohibited high-impact decisions without required safeguards.
Transparency
Where appropriate, customers should inform individuals when they are interacting with an AI-powered system.
Customers are responsible for complying with applicable disclosure and transparency requirements.
Monitoring and Safety
NaviCS AI may use technical and operational controls to detect abuse, protect the platform, investigate security events, and enforce applicable policies.