Reversibility
The degree to which a completed payment can be undone or recovered after it has been sent, ranging from a strong, legally enforceable right to a real recovery process (as with a properly reported unauthorized ACH transfer) down to a complete technical and legal impossibility (as with a confirmed cryptocurrency transaction or cash handed to a courier).
Defined in 5 GAGE programs, which carry 15 distinct definitions of it. The wording above is taught in Senior AI Scam Defense.
How each discipline defines it
The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.
The degree to which a completed payment can be undone or recovered after it has been sent, ranging from a strong, legally enforceable right to a real recovery process (as with a properly reported unauthorized ACH transfer) down to a complete technical and legal impossibility (as with a confirmed cryptocurrency transaction or cash handed to a courier).
How easily an action the agent takes can be undone. A draft is reversible (discard it); a sent message, posted statement, executed payment, or deleted record may be irreversible. Often the deciding factor in a delegation, because an irreversible action has no undo when the agent is wrong.
Whether a wrong action, once taken, can be caught and undone before it causes further harm. A low-reversibility action combined with no human checkpoint is the highest-risk combination the actions-versus-outputs test can surface, and is the exact shape of the Replit database deletion.
The degree to which an action's effects can be undone. Full (a deletable draft), partial (a clawback-able refund), or none (a sent email, an external wire, a physical or rights-affecting act). Lower reversibility pushes toward in-the-loop.
the property, delivered by incremental patterns like the strangler fig, that a failed change can be undone by routing a single migrated capability back to the still-running old system, rather than losing the whole business at once.
Where it is taught
The exact lessons this term appears in. The first module of every program is free with a free account.
- When NOT to Use AI: Decision Framework and Technical Debt Risk · Practical AI Workflow Design and Prompt Engineering, Certified AI Practitioner: Workplace Foundations
- Risk Assessment in AI Decisions · Critical Thinking and Context Engineering, Certified AI Practitioner: Workplace Foundations
- Working Alongside AI Agents: Collaboration Frameworks · Agentic AI and Workforce Integration, Certified AI Practitioner: Workplace Foundations
- When to Delegate to AI Agents vs. Handle Tasks Yourself · Agentic AI and Workforce Integration, Certified AI Practitioner: Workplace Foundations
- Strategic AI Decision-Making Frameworks · Bonus: AI Leadership Accelerator, Certified AI Practitioner: Workplace Foundations
- Why Agentic Breaks Generative Governance · Taking the Controls, Certified Agentic AI Governance Professional (CAAGP)
- The Four Levels of Human Involvement · Capability and Autonomy, Certified Agentic AI Governance Professional (CAAGP)
- Risk Tiering in Practice · Capability and Autonomy, Certified Agentic AI Governance Professional (CAAGP)
- Checkpoint Design · Meaningful Human Accountability, Certified Agentic AI Governance Professional (CAAGP)
- Dossier Workshop D: Accountability Map · Meaningful Human Accountability, Certified Agentic AI Governance Professional (CAAGP)
Where it shows up at work
A definition is where Reversibility starts. The lessons above teach it inside the competencies below, which are what job descriptions ask for and what the GAGE credential records. Every link here is drawn from the same careers map that builds a role's readiness check.
Knows when to trust, verify, escalate or override an AI recommendation, and stays accountable for the decision.
5 lessons on this page teach it as a primary outcome.
Defines who reviews AI outputs, what they check, when they can override, and how to keep review from becoming a rubber stamp.
4 lessons on this page teach it as a primary outcome.
Briefs a model like a good manager: context, constraints, examples, verification, and knows when to stop delegating.
3 lessons on this page teach it as a primary outcome.
Roles that hire for it
Each role page shows the full competency set, what the seat decides, and a readiness check against your own evidence.
- Operations Manager, AI-readyneeds it at depth
- AI Literacy Leadneeds it at depth
- AI Product Manager, responsible product editionneeds it at depth · 3 open postings this week
- Ethical AI Specialistneeds working knowledge · 5 open postings this week
- VP of AI Governance and Ethicsneeds working knowledge
- Autonomous Systems Governance Lead, robotics and physical AI editionneeds working knowledge
How job descriptions say it
Postings rarely use the textbook word. These are the phrases the careers map reads as asking for the competencies above.
Where the frameworks place it: EU AI Act Article 14; NIST AI RMF GOVERN; ISO/IEC 42001 Annex A.
One interview question about it
A model recommends a decision that affects a person. What do you check before you act on it, and when do you overrule it?
Asked of candidates for judgment when ai supports a decision. Answering it in your own words is exactly how a GAGE topic is passed, and the credential records the pass where an employer can check it.
If you come from another field
Some backgrounds already carry this competency, and the readiness check gives credit for it.
- Operations, claims, customer service and administration carries it as a claim to back up: Deciding when to escalate a case is judgment under uncertainty.
- A computer science or software engineering degree carries it as work you can show: Prompting and tool workflows are already in your daily work.
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.