AI Output and Review Policy
Effective date: 17 April 2026
1. DGS outputs are decision-support materials
DGS produces mission artifacts, reports, maps, proof objects, summaries, recommendations, and other generated materials. These outputs are intended to support professional review and execution. They are not automatic truth, legal advice, financial advice, medical advice, engineering certification, safety approval, regulatory approval, or a guarantee of outcome.
2. Human review is required
You are responsible for reviewing, testing, validating, and approving outputs before relying on them. For regulated, safety-critical, nuclear, legal, financial, medical, engineering, compliance, employment, credit, housing, insurance, or other high-impact uses, review by qualified professionals is required before action.
3. No guaranteed correctness or innovation result
DGS may synthesize strong structures and proof objects, but outputs can be wrong, incomplete, outdated, misleading, unsuitable, or unsafe for a specific use. DGS does not guarantee accuracy, novelty, patentability, compliance, commercial value, investment outcome, safety, technical feasibility, regulatory acceptance, or business success.
4. User responsibility for real-world action
You are responsible for decisions, publications, filings, deployments, purchases, investments, operational changes, engineering changes, compliance actions, and other real-world steps taken using DGS outputs. You must independently confirm source material, assumptions, calculations, dates, legal requirements, and implementation constraints.
5. Public proof and restricted internals
Public proof objects may show selected examples of DGS capability. They do not expose all internal methods, restricted engine topology, private evidence, or customer workspace materials. Access to restricted systems remains subject to separate approval.