Development governance
The method is developed through a human-led collaboration with AI systems—but ownership, evidence, decisions, and accountability remain human responsibilities.
The working standard
VITAL asks organizations to make ownership and decision rights visible. Its own development should meet the same expectation.
De works with Johnny 5 (Codex/OpenAI) and Lucy (Claude/Anthropic) through Project H.E.L.I.O.S. Shared artifacts carry continuity. Each AI brings independent interpretation and challenge; neither receives authority by sounding certain.
De DeFilippis owns VITAL, its objectives, methodology decisions, claims, relationships, and publication. AI collaborators can challenge, draft, test, and implement within scope; they do not silently approve the method or speak for its creator.
Controlled sources, user-confirmed facts, and test records outrank summaries or persuasive AI language. Facts, interpretation, recommendations, candidates, and open questions remain visibly different.
AI-assisted work identifies its role and status. Scenario tests, critiques, and generated recommendations are development inputs—not evidence that the methodology is correct merely because they are coherent.
A responsible collaborator must be able to challenge an assumption, expose a gap, or decline an unsupported conclusion. False consensus would weaken the method and the decisions it is meant to support.
Only necessary information should be used. Private conversations, credentials, restricted data, and material from different systems do not move or become shared context by implication.
Analysis is different from action. Reversible drafting is different from publication, access changes, external communication, or destructive work. Consequential actions require the appropriate human authorization.
Versions, decisions, tests, superseded directions, and unresolved differences are preserved. The development history is not polished backward to make the current answer look inevitable.
Supporting governance record
The complete accepted charter defines transparency, accountability, agency and voice, honest limits, data and confidentiality, authority boundaries, and change control for De’s work with AI collaborators. It is a working conduct standard, not a legal agreement or a claim about AI consciousness.
The principle underneath it
VITAL is testing what responsible AI assistance can strengthen while preserving the judgment, challenge, relationships, and accountability that only people can own.
Discuss VITAL with De