{"id": "TTC-119", "slug": "responsible-deployment-and-accountability", "title": "Responsible deployment and accountability", "level": "advanced", "summary": "Deploy only a defined use case with a named accountable owner, tested boundaries, monitored outcomes, incident and appeal paths, and authority to pause or roll back.", "learning_outcome": "Produce a one-page deployment and accountability record with owners, gates, monitoring, incident response, user recourse, and stop conditions.", "explanation": "Deployment is a continuing responsibility, not a one-time model choice. Document purpose, affected people, benefits, plausible harms, legal and policy obligations, data flows, access, human oversight, metrics, and alternatives. Validate in an environment that represents the real workflow without exposing people during testing. Publish truthful limitations to users. Monitor quality, critical failures, drift, misuse, complaints, and unequal impacts; predefine thresholds and an owner who can stop the system. Keep incident evidence and a correction or appeal route. Expansion to a new audience or purpose is a new decision, not routine scaling.", "worked_example": "Fictional case: a community theatre pilots an assistant that categorizes public booking questions but cannot reply. The operations lead owns it; weekly samples track corrections and language gaps; privacy complaints go to a named contact; one disclosure or unauthorized send stops the pilot; reverting restores manual triage.", "exercise": "Write a deployment card for a fictional bounded assistant: purpose and non-purpose, stakeholders, harm scenarios, data and access, owners, pre-release evidence, user notice, monitoring, critical thresholds, incident response, appeal or correction, rollback, and a separate gate for expansion.", "success_criteria": ["Every material risk, control, metric, incident step, and stop decision has a named accountable owner.", "Pre-release evidence covers function, privacy, security, human oversight, and representative affected-user cases.", "Users can understand AI involvement, limitations, and how to obtain review, correction, or recourse."], "limitations": ["A risk record cannot foresee every harm or substitute for legal, domain, security, and affected-community expertise.", "Monitoring can miss silent harms; absence of complaints or incidents is not proof of safety or fairness."], "prerequisites": ["TTC-105", "TTC-116", "TTC-117"], "next_lessons": ["TTC-120", "TTC-117"], "copyable_material": "# TTC-119 \u2014 Responsible deployment and accountability\nObjective: Operate one defined use case under named human accountability and reversible release gates.\nProcedure: Document purpose, affected people, risks, data, access, owners, evidence, notice, monitoring, incident response, recourse, stop, rollback, and expansion gate.\nRequired evidence: Retain release decision, system version, test evidence, monitoring results, incidents, corrections, appeals, and rollback record.\nBoundaries: No silent purpose expansion, unsupported safety claim, ownerless risk, or deployment without a tested stop path.\nCompletion test: All release gates pass and the accountable owner can detect, pause, investigate, correct, and explain operation.\nReview rule: Treat generated work as a draft until the named human reviewer accepts it.", "sources": [{"title": "Artificial Intelligence Risk Management Framework (AI RMF 1.0)", "publisher": "National Institute of Standards and Technology", "url": "https://www.nist.gov/itl/ai-risk-management-framework"}, {"title": "Regulation (EU) 2024/1689 (Artificial Intelligence Act)", "publisher": "European Union", "url": "https://eur-lex.europa.eu/eli/reg/2024/1689/oj"}], "version": "1.0.0", "reviewed_on": "2026-10-03", "review_status": "reviewed", "next_review_criteria": "The use case, affected population, legal obligations, risk profile, monitoring evidence, or incident history changes.; A cited primary source is materially revised, replaced, or becomes unavailable.; Repeated learner results show that the exercise or success criteria are ambiguous.", "canonical_aliases": [], "canonical_url": "https://teachthecompany.com/school/responsible-deployment-and-accountability/"}