4Responsible.AI Knowledge Graph — governed by Propel / Protect / Preserve / Bound:
| FOUR CORE PrincipleS | Standard components TO ACHIEVE NORMATIVE REQUIREMENTS |
|---|---|
| Propel | Generative Learning AI Skills Taxonomy Cognitive Lean Loop |
| Protect | Human Agency Human Agency Rails PACE |
| Preserve | Atrophy Risk Assessment Human Skill Encoding Generative Coevolution |
| Bound | Agentic Delegation Process PIAF Human Coding / CODE |
How the 4R principles map to external frameworks — NIST, DOL / AI Literacy, UNESCO / AILit, and enterprise governance:
| 4R Principle | NIST | DOL / Literacy | UNESCO / AILit | Enterprise Governance |
|---|---|---|---|---|
| Propel | Map / Manage opportunity | Build AI knowledge and application capability | Engage / Create | Enable responsible innovation |
| Protect | Govern / human oversight | Critical evaluation | Human-centered mindset / Manage | Accountability / human oversight |
| Preserve | Human impact / sociotechnical risk | Adaptability and durable worker capability | Human agency / critical thinking | Relative gap in most enterprise frameworks |
| Bound | Govern / Map / Measure / Manage | Responsible-use knowledge | Ethics / Manage AI | Controls, testing, monitoring, security |
| Reorient / Coevolve | Continuous risk management | Continuous learning | Progressive competency | Lifecycle monitoring |
How 4Responsible.AI compares with adjacent AI risk, literacy, and governance frameworks:
| Framework | Primary question | Primary unit of concern | Core approach | AI risk & governance | Workforce adoption | Distinctive strength | Relationship to 4Responsible.AI |
|---|---|---|---|---|---|---|---|
| 4Responsible.AI | How should people, teams and organizations evolve as AI becomes more capable? | Individual → Team → Organization | Propel → Protect → Preserve → Bound, supported by PIAF, PACE, ADP, Atrophy Risk, Human Coding and maturity levels | Medium with 7-Normative Requirements, Establishing Standard* | Very High | Connects adoption, delegation, human skill preservation and organizational evolution | Workforce/human-agency operating layer. Establishing metrics and evidence* |
| NIST AI RMF | How do we identify and manage AI risk? | AI system + organization | Govern → Map → Measure → Manage | Very High | Medium-Low | Rigorous, technology-neutral risk-management architecture | 4R can become the human/workforce implementation companion to NIST |
| UNESCO AI Competency Framework | What should people know and be able to do with AI responsibly? | Learner / educator | Human-centered mindset, ethics, techniques/applications, system design; progression through Understand → Apply → Create | Medium | Medium | Explicit human-centered competencies and developmental progression | Strong philosophical alignment; 4R extends competency into workplace behavior and organizational design |
| U.S. DOL AI Literacy Framework | What baseline AI literacy does the workforce need? | Worker / training ecosystem | Five foundational content areas + seven delivery principles | Medium | High | National workforce-training orientation and adaptable program design | 4R could provide the post-literacy adoption pathway after foundational DOL literacy |
| AILit Framework | What knowledge, skills and attitudes enable meaningful AI participation? | Primarily learners | Engage with AI → Create with AI → Manage AI → Shape AI | Medium | Medium-Low | Integrates knowledge, skills, attitudes, expectations and scenarios | 4R extends "Manage/Shape AI" into delegation, work redesign and preservation |
| Microsoft Responsible AI | How do we design, deploy and operate trustworthy AI? | AI product/system + enterprise | Fairness, reliability/safety, privacy/security, inclusiveness, transparency, accountability + Responsible AI Standard | Very High | Medium | Converts principles into engineering and governance requirements | 4R fills a workforce-development gap: what employees do differently because AI exists |
| Google Responsible AI / SAIF | How do we innovate while controlling AI lifecycle and security risks? | Models, applications, infrastructure | Bold innovation + responsible development + collaborative progress; lifecycle testing, safeguards and SAIF controls | Very High | Medium-Low | Deep technical lifecycle governance, security and risk controls | Google governs the AI asset; 4R can govern the human-AI work relationship |
| IBM watsonx.governance | How can enterprises continuously govern models, applications and agents? | AI asset portfolio | Inventory → factsheets → evaluation → monitoring → policies/controls → lifecycle governance | Very High | Medium | Operational governance tooling, monitoring, auditability and multi-model lifecycle control | IBM provides the technical governance plane; 4R provides the human/workforce governance plane |