AI & Automation Policy
This policy sets minimum governance rules for AI agents and automated workflows designed, supplied or operated by HITCHNOVA.
Probabilistic Outputs
AI systems may produce incomplete, inaccurate, outdated or unsuitable outputs. Appropriate human review is required, especially where an output could affect people, money, legal obligations, safety or material business decisions.
No Independent Executive Authority
AI agents must not independently approve payments, sign contracts, make final executive decisions, override governance controls or create binding obligations unless a narrow, pre-approved automated action has been lawfully configured with appropriate controls.
Disclosure
Where an end user materially interacts with an AI agent, the client should provide appropriate notice that the interaction is automated or AI-assisted where required by law, contract or good practice.
High-Risk Uses
HITCHNOVA solutions must not be represented as a substitute for licensed medical diagnosis, final legal advice, regulated credit decisions, final employment decisions or other high-impact professional determinations without qualified human oversight and any required approvals.
Voice Agents
Where voice agents, call recording or transcription are used, the client is responsible for ensuring required notices, permissions and consents are provided before activation.
Data and Permissions
Only the minimum information and permissions necessary for the approved task should be used. Sensitive actions should use approval gates, scoped credentials, re-authentication, rate limits, audit logging or other controls proportionate to risk.
Untrusted Content
External documents, webpages, emails, messages and user content must be treated as potentially untrusted. AI systems should not follow embedded instructions that conflict with approved system rules, security controls or client authority boundaries.
Auditability and Kill Switch
Material automated actions should be logged where technically feasible. Production AI systems should support suspension, credential revocation or emergency disablement where technically appropriate.
Contact
Questions may be sent to hitchnova.official@gmail.com until dedicated domain addresses are activated and tested.
Versioning
Version: 1.0. Effective date: 19 September 2026.