📇 ACCT 3233/7233 Exam Notecards — Print on 4×6 index cards (landscape)
| Domain | Focus |
|---|---|
| I Purpose | Why we exist |
| II Ethics | How we behave |
| III Governing | Board & CAE relationship |
| IV Managing | Strategy & resources |
| V Performing | The actual work |
Goal: create, protect, and sustain value
| P# | Principle | Philosopher |
|---|---|---|
| P1 | Integrity | Kant — categorical imperative |
| P2 | Objectivity | Husserl — epoché (bracket assumptions) |
| P3 | Competency | Aristotle — phronesis (practical wisdom) |
| P4 | Due Prof. Care | Hume — can't test everything |
| P5 | Confidentiality | Breach once → never candid again |
Defines "Essential Conditions" boards must provide.
⚠️ New term: "conclusion" replaces "opinion"
Most failures = planning failures. No surprises in final report = you've talked to mgmt first.
| Line | Who | Role |
|---|---|---|
| 1st | Management | Owns risk & controls day-to-day |
| 2nd | Risk & Compliance | Supports, monitors, provides expertise |
| 3rd | Internal Audit | Independent assurance → board |
| Gov | Board | Accountability to stakeholders |
IA can rely on 2nd line but must independently verify before reporting conclusions.
| Assurance | Advisory | |
|---|---|---|
| Parties | 3 (auditor, auditee, stakeholders) | 2 (auditor + client) |
| Scope | Auditor sets | Client sets |
| Output | Formal conclusion | Recommendations |
| Posture | Skepticism | Collaboration |
⚠️ Advisory on a process may impair objectivity for future assurance (Std 7.2)
Mandatory when in scope. Effective 12 months after issuance. Must document which apply + rationale for exclusions.
| Topic | Key Focus |
|---|---|
| Cybersecurity | Governance, risk, controls, incident response |
| IT Governance | Tech strategy, performance, risk |
| Privacy Risk Mgmt | GDPR, CCPA, privacy by design |
| Third-Party Mgmt | Vendor risk, due diligence, monitoring |
| Sustainability/ESG | Climate, supply chain, DEI metrics |
"The unexamined organization is not worth working for." — with apologies to Socrates
| Objective | Protects against | Threats |
|---|---|---|
| Confidentiality | Unauthorized disclosure | Breaches, eavesdropping, unauthorized access |
| Integrity | Unauthorized modification | Tampering, malware, MITM attacks |
| Availability | Disruption | DDoS, ransomware, system failures |
Every control should map to ≥1 CIA objective. If it doesn't — question why it exists.
Market volatility doesn't try to defeat your hedging strategy. Cyber attackers do.
Risk-based prioritization: What is most critical? Most exposed? Would hurt most if compromised?
| Type | Examples |
|---|---|
| Preventive | Firewalls, access controls, encryption, patching, security training |
| Detective | SIEM, IDS/IPS, log monitoring, anomaly detection, threat hunting |
| Corrective | Incident response, backup/restore, disaster recovery, forensics |
Layers: Network → Endpoint → Application → Data → Human → Governance
Human layer = often weakest AND strongest. Vigilant employees catch what tools miss.
| # | Pathology | Key Example / Audit Q |
|---|---|---|
| 1 | Ownership Gap — unowned systems don't get patched | Target 2013: HVAC vendor. "Show me who owns this system and prove they know it." |
| 2 | Security vs. Business — speed vs. control; CISO reporting line matters | "Show me a project where security delayed launch. What happened?" |
| 3 | Compliance Trap — compliance ≠ security | Equifax, Target, Capital One all PCI compliant when breached. Attackers don't check compliance status. |
| 4 | Learned Helplessness — "if they want in they'll get in" fatalism | Verizon DBIR: stolen credentials + phishing + unpatched systems cause MOST breaches. Fundamentals > exotic attacks. |
"Did the control work when it should?"
Testing the happy path
Verify approvals, timelines, documentation
"Did the control BLOCK what it should?"
Testing the guardrails
Can terminated users log in? Self-approve? SoD blocked?
"If everyone followed this perfectly, would it prevent/detect the risk?"
| Design Failure | OE Failure | |
|---|---|---|
| What | Control CAN'T work | Control DIDN'T work |
| Why | Flawed blueprint: policy gap, wrong approvers, no SLA | Good design, bad execution: approval skipped, SLA breached |
| Remedy | Process/policy redesign | Training, monitoring, automation |
Golden Rule: If you can't point to a specific exhibit/row/doc for every fact → not documented properly.
| Method | Use When | Limitation |
|---|---|---|
| Random | Large, homogeneous populations | May miss high-risk by chance |
| Stratified | Heterogeneous pop. with risk strata | Must understand risk factors |
| Haphazard | Small populations; preliminary testing | NOT statistically valid; unconscious bias |
| Census (100%) | Critical controls; small populations; high-risk | Time-consuming |
Reviewer's Test: Could another auditor recreate your exact sample?
| Frequency | Low | Mod | High |
|---|---|---|---|
| Annual (1) | 1 | 1 | 1 |
| Quarterly (4) | 2 | 2–3 | 4 |
| Monthly (12) | 2–3 | 4–6 | 8–12 |
| Weekly (52) | 5–9 | 10–15 | 20–25 |
| Daily (250+) | 10–15 | 20–30 | 40–60 |
| Per occurrence | 25 | 40 | 60 |
| Population | Risk | Method | Sample |
|---|---|---|---|
| Privileged users | Critical | Census | 100% |
| SoD conflicts | Critical | Census | 100% |
| New provisioning | High | Stratified | 60 |
| Terminations | High | Stratified | 52 |
| Modifications | Medium | Random | 45 |
| Rating | Definition | Consequence |
|---|---|---|
| Material Weakness | Reasonable possibility material misstatement NOT prevented/detected | Adverse ICFR opinion |
| Sig. Deficiency | Less than MW but merits oversight attention | Report to Audit Committee |
| Ctrl Deficiency | Design/operation gap; may not reach SD/MW threshold | Management letter |
| External (Mgmt Letter) | Internal (IA Finding) | |
|---|---|---|
| Structure | Condition → Criteria → Cause → Effect → Rec. | Title → Background → Observation → Risk → Rec. + Owner + Due Date + Mgmt Response |
| Tone | Formal, regulatory | Collaborative, advisory |
| Specificity | "Certain users" (vague) | Named individuals, specific data points |
| Audience | Audit Committee / regulators | Management + CAE + Board |
| Rule | What It Means |
|---|---|
| 1. Lead with headline | Exception rate goes first — "2 of 60 (3.3%) had exceptions" |
| 2. Quantify everything | Never "several" or "many" → always "3 of 127 (2.4%)" |
| 3. Fact vs. judgment | "16 hrs elapsed" = fact · "deficiency" = judgment · label each clearly |
| 4. Define your population | State the universe: "Of 412 new provisioning events during the test period…" |
| Rule | What It Means |
|---|---|
| 5. Show your math | "16.75 hrs (Fri 5pm → Sat 9:45am) vs. 4-hr SLA" — never assert without the calculation |
| 6. Compensating controls | Acknowledge them: "No unauthorized access confirmed via sign-in logs" = credibility |
| 7. Actionable recs | Not "improve controls" → "implement real-time Workday-to-Entra sync by Q2" |
| 8. Write for the skeptic | Pre-answer every "how do you know?" — assume the reviewer challenges every assertion |
Totalitarian power sustains itself not through constant overt violence, but through the daily participation of ordinary people in lies.
Violence needs lies to maintain respectability. Lies need violence to be enforced. They are symbiotic.
But violence cannot act on every shoulder every day. All it demands: obedience to lies and daily participation in lies.
The "simplest and most accessible key to liberation": personal non-participation in lies.
Not revolution. Not marching into squares. Not hunger strikes. Simply: refuse to say what you do not think.
"The most moderate of all paths of resistance" — easier than Gandhi's civil disobedience. The flames will not touch your body.
| Concept | Key Point |
|---|---|
| Fear hierarchy | People don't fear nuclear war. They fear acts of civil courage — losing white bread, heating gas, Moscow registration. |
| Spiritual independence vs. servitude | Every day: either truth (spiritual independence) or lies (spiritual servitude). No neutral ground. |
| The code of conduct | Won't sign/write phrases distorting truth; won't cite out of context to please; will walk out of meetings where lies are told; won't buy publications that distort facts. |
| Czechoslovakia | Unarmed people with worthy hearts can stand up to tanks. If we are thousands, they cannot stop us all. |
| Pushkin closing | "What use to the herds the gifts of freedom? The scourge and a yoke with tinkling bells." Those who choose comfort over truth deserve their suffocation. |
Self-respect is an internal standard — entirely independent of others' approval or reputation. Its source is character: the willingness to accept responsibility for one's own life.
Self-deception is the most difficult deception. No winning smiles or lists of good intentions work in the well-lit back alley where one keeps assignations with oneself.
| Concept | What It Means |
|---|---|
| Self-respect ≠ approval | Has nothing to do with others' approval — who are, after all, deceived easily enough. Rhett Butler: reputation is what people with courage can do without. |
| Character | "The willingness to accept responsibility for one's own life" — the source of self-respect; sometimes loses ground to more negotiable virtues. |
| Interminable documentary | Without self-respect: unwilling audience to endless replay of one's failings, broken promises, gifts wasted through sloth/cowardice. |
| Concept | What It Means |
|---|---|
| Jordan Baker vs. Julian English | Baker (dishonest, Great Gatsby) HAD self-respect — made her own peace. English (suicidal, Appointment in Samarra) did NOT. Self-respect ≠ conventional virtue. |
| Discipline as habit | Self-respect is "a discipline, a habit of mind that can be developed, trained, coaxed forth." Small disciplines represent larger ones — reminders of who we are. |
| Alienation from self | Advanced stage: can't say no without drowning in self-reproach. Eventually runs away to find oneself and finds no one home. |
| Intrinsic worth | To have it = potentially have everything: ability to discriminate, love, remain indifferent. To lack it = locked within oneself, paradoxically incapable of either love or indifference. |
| Role | Who They Are | What Happens |
|---|---|---|
| Teacher | Naïve real subject | Administers shocks; the true focus of the study |
| Learner | Actor / confederate | Receives no real shock; pretends to suffer |
| Experimenter | Authority figure | Gives prods: "The experiment requires you continue" |
Shock generator: 15–450 volts, labeled Slight → Danger: Severe Shock → XXX. Each subject first receives a real 45-volt sample to confirm authenticity.
Predicted: Psychiatrists forecast only a pathological fringe (~1 in 1,000) would reach 450 volts.
Actual: 25 of 40 subjects (62.5%) obeyed to the maximum 450-volt shock. Results replicated across Yale students, New Haven community members, and internationally (Munich: 85% obedient).
| Variation | Effect |
|---|---|
| Orders given by telephone (not in person) | Obedience drops to ~⅓ of normal |
| Two experimenters give conflicting orders | No shocks delivered past disagreement point |
| Two peers disobey first | 36 of 40 subjects joined and refused |
| Subject not ordered directly (subsidiary task) | 37 of 40 continued to highest level — "I only administered the test" |
| Concept | Definition / Key Point |
|---|---|
| Banality of Evil | Arendt: Eichmann not a sadistic monster — an uninspired bureaucrat. Milgram confirms: ordinary people, not sadists, deliver 450 volts out of duty. |
| Diffusion of responsibility | Subject who only "administers the test" credits responsibility to the person who "actually pulled the switch." Classic in complex organizations. |
| Physical presence of authority | Experimenter in room → high obedience. Phone orders → obedience drops sharply. |
| Subjects' motivation | NOT sadism — they derive satisfaction from pleasing authority ("proud of doing a good job"), not from inflicting pain. |
| Solzhenitsyn | Didion | Milgram | IIA Principle |
|---|---|---|---|
| Non-participation in lies | Character: accept responsibility | Defying the agentic state | P1 Integrity / Std 1.1 |
| Fear of civil courage | Seeking others' approval | Obedience to authority | P2 Objectivity — independence |
| Spiritual independence | Intrinsic worth | Personal moral agency | P1 + P2 together |
| "Refuse to say what you don't think" | "Discipline, habit of mind" | Peer who disobeys first | P4 Due Professional Care |
| Virus needs living host | Interminable documentary | Diffusion of responsibility | Fraud depends on auditor silence |
Softening a finding = daily participation in lies
Audit manager who removes a finding = the living host the lie needs
Seeking management approval for findings = opposite of self-respect
Confirmation bias = self-deception (most difficult deception)
"I just followed the partner's direction" = agentic state; no absolution
Speak-up culture = Milgram's peer disobedience variation — one voice changes the room
"The unexamined organization is not worth working for." — with apologies to Socrates
| IIA Ethics Principles & Philosophers | |
|---|---|
| P1 Integrity | Kant's categorical imperative — universal law |
| P2 Objectivity | Husserl's epoché — bracket assumptions; beware confirmation bias |
| P3 Competency | Aristotle's phronesis — practical wisdom; use specialists |
| P4 Due Prof. Care | Hume — can't test everything; sample intelligently |
| P5 Confidentiality | Breach once = never candid again |
| Four C's · Engagement · Key Terms | |
|---|---|
| Four C's | Condition · Criteria · Cause · Consequence |
| Effort split | Planning 20-30% · Fieldwork 40-50% · Reporting 15-20% |
| EQA | Every 5 years · ≥1 active CIA required (Std 12.4) |
| Conclusion | New 2024 term — replaces "opinion" |
| Topical Reqs | Mandatory · Effective 12 months after issuance |
| CIA Triad | |
|---|---|
| Confidentiality | Unauthorized disclosure → breaches, eavesdropping |
| Integrity | Unauthorized modification → tampering, MITM, malware |
| Availability | Disruption → DDoS, ransomware, system failures |
| ITGC Key Facts | |
|---|---|
| Positive test | Happy path — did it work when it should? |
| Negative test | Guardrails — did it block what it should? |
| TOD | Blueprint — can the control work? (Precision/Completeness/Authority) |
| Design fail | Can't work → redesign (don't prescribe training!) |
| OE fail | Didn't work → training, monitoring, automation |
| MW | Reasonable possibility material misstatement → adverse ICFR opinion |
| Sig. Deficiency | Less than MW → report to Audit Committee |
| Census | 100% — always for privileged users, SoD conflicts |
| Daily/High | 40–60 samples (AICPA/PCAOB) |
Good luck on the exam!