Review a compact APMG AI Project Governance Framework Foundation cheat sheet for governance roles, controls, responsible AI, lifecycle gates, assurance, and metrics before PM Mastery practice.
Use this AIPGF Foundation cheat sheet to review the governance controls behind AI project decisions. Foundation questions usually test recognition: who is accountable, what control is missing, which lifecycle checkpoint applies, and how responsible AI principles become reviewable project behavior.
| Item | AIPGF Foundation cue |
|---|---|
| Provider | APMG International |
| Exam | AI Project Governance Framework (AIPGF) Foundation |
| Format focus | 40 questions in 40 minutes |
| Practice behavior | recognize the governance control, role, principle, lifecycle checkpoint, or assurance measure that fits the AI project scenario |
| PM Mastery status | live practice available |
| Area | What to know | Common trap |
|---|---|---|
| Framework purpose | accountability, controls, risk treatment, assurance, and responsible AI behavior | treating governance as paperwork after the AI work is done |
| AI context | project objectives, organizational risk, data use, vendors, stakeholders, and operating impact | assuming AI is only a technical delivery issue |
| Controls | approvals, evidence, human review, prompt/output logging, escalation, and change control | using a control name without matching it to the risk |
| Roles | accountable owner, project manager, reviewer, approver, vendor, user, and assurance function | letting responsibility bounce between functions |
| Responsible AI | transparency, fairness, privacy, security, explainability, and human oversight | stating principles without operational safeguards |
| Culture and behavior | openness, challenge, safe reporting, learning, and ethical use | blaming users when governance is unclear |
| Lifecycle governance | initiate, design, build, test, release, operate, monitor, and improve | allowing a pilot to bypass stage controls because it is experimental |
| Assurance and metrics | measures, thresholds, review cadence, incidents, lessons, and improvement | measuring tool usage instead of control effectiveness |
After each AIPGF Foundation set, classify misses by control type: role/accountability, lifecycle gate, responsible AI principle, vendor/data risk, human review, or assurance metric. If you can name the principle but not the control, drill scenario questions before repeating definitions.