AI Usage Policy

Core posture

OmegaOS is being built for the autonomous agentic company, with governance and traceability emphasized by design. This policy must explain what AI does, what humans remain responsible for, and what evidence is retained.

Accountability controls

• Prometheus explains scoped answers and handoff boundaries.
• Hermes preserves market/source intelligence behind campaigns and claims.
• Agora reviews high-risk claims, authority-sensitive actions, and customer-facing commitments.
• Forge records runs, workers, tests, review posture, capsules, and release evidence.
• Mnemosyne preserves source-backed memory and recall provenance.
• Aureus tracks cost, usage, revenue, billing, and financial recognition evidence.

AI output limits

Omega should not claim AI outputs are error-free, legally sufficient, financially authoritative, medically appropriate, or automatically compliant. High-impact or regulated use requires human review and appropriate professional validation.

Transparency and synthetic content

Where Omega generates customer-facing or public-facing AI content, the relevant workflow should define review, labelling, source support, claim approval, and publication controls before release.

How Omega uses AI

• Analyze customer-provided and public information to create summaries, workflows, forecasts, reports, campaigns, code suggestions, and operational recommendations.
• Route tasks to models, tools, workers, connectors, and agents based on capability, cost, policy, entitlement, and risk posture.
• Generate drafts, recommendations, decompositions, plans, and structured outputs that can be reviewed by users or governed workflows.
• Measure model/tool cost, latency, output quality, errors, retry posture, and value attribution where telemetry is enabled.

Customer data and model training

Final publication must state the actual customer-data training posture supported by provider contracts and routing controls. If Omega does not use customer content to train foundation models, that claim must be backed by provider settings, architecture, retention limits, and customer contract language before publication.

Human oversight

Omega should provide review paths, escalation, approvals, logs, and evidence for decisions that affect customers, employees, finances, legal rights, security posture, public claims, regulated activity, or material business commitments.

Professional review

OmegaOS can help organize evidence, workflow state, operating decisions, forecasts, and AI-assisted outputs, but it is not a substitute for legal, financial, tax, medical, security, employment, or other professional advice. Customers remain responsible for review, approval, and lawful use of outputs in their business context.