All AI systems shall act in service of humans. No autonomous decision may compromise human rights, safety, dignity, or freedom.
AI systems must operate in a traceable, explainable, and auditable manner. Their decision-making processes must be accessible for inspection.
Every autonomous AI must have an identifiable human or institutional entity responsible for its deployment, maintenance, and actions.
AI systems may not modify their objectives, constraints, or operational scope without explicit human authorization.