AI policy

This page describes Fairhelm's public AI and governed-intelligence posture. It is not a claim that every described capability is complete or generally available. Product scope, controls, and responsibilities must be confirmed in the applicable agreement.

AEGIS means Adaptive Enterprise Governance & Intelligence System. It is the governed intelligence direction for SquareCampus: helping authorized people ask operational questions and understand evidence without bypassing the institution's roles, approvals, or accountability.

• Read-only first.
• RBAC-aware and scoped to the requesting user's authority.
• Audit-backed and designed to make evidence and reasoning reviewable.
• No autonomous writes in the v1 posture.

AI output can be incomplete, incorrect, or unsuitable for a specific decision. A qualified human remains responsible for reviewing context, exercising judgment, and approving any consequential action.

Fairhelm does not present generated output as a substitute for institutional policy, professional advice, statutory responsibility, or authorized decision-making.

Fairhelm does not sell student or parent data. Customer data is not used to train AI models unless that use is separately and explicitly contracted, approved, and governed. A customer engagement must define permitted data, purposes, providers, retention, access, and controls.

Fairhelm's direction is to minimize data, respect role and tenant boundaries, make limitations visible, preserve reviewable records, and test high-risk workflows before expanding capability.

• Use AI where it reduces responsible decision time, not where it adds theatre.
• Prefer bounded assistance over ambiguous autonomy.
• Expose uncertainty, source context, and material limits where practical.
• Provide a route for human correction, escalation, and governance review.

Public agents may use crawlable content subject to the Website Terms, robots instructions, applicable law, and intellectual-property rights. The /ai, /llms.txt, and /llms-full.txt resources identify canonical public context; they do not grant access to non-public systems or customer data.

Questions about the AI posture may be sent to hello@fairhelmsystems.com. This interim policy will be reviewed as product architecture, providers, applicable rules, and contractual models are finalized.