Foundry Academy · AI Workflow Training · Lesson 1 of 6

Task and risk classification

Decide whether and how AI may assist a task by evaluating harm, data sensitivity, reversibility, expertise, scale, and available controls.

Start the lesson

Your learning work, on this device

No signup, cloud storage, cross-device sync or verified completion. Saving is optional. This browser profile is shared with anyone who can use it; private mode, browser cleanup or storage limits may remove work. Use only fictional or non-sensitive material. Export a copy before relying on this device.

Not saved. Worksheets, answers and practice notes currently last only in this tab.

Practice markers are self-reported, never credentials.

01 · Explanation

Task and risk classification

Objective: Decide whether and how AI may assist a task by evaluating harm, data sensitivity, reversibility, expertise, scale, and available controls.

Begin with the task, not the tool. Describe the current input, output, user, decision, frequency, owner, and failure consequence. Separate assistance from authority: drafting a meeting summary is different from approving a loan, diagnosing a condition, selecting a candidate, or changing a supplier. Score potential impact on people, finances, safety, rights, reputation, security, and operations. Then consider data sensitivity, reversibility, need for licensed or specialist judgment, likelihood that users will overtrust the output, and scale. A frequent low-severity error can create more harm than an obvious one-off failure.

Choose a treatment only after classification. Low-risk tasks may use AI with ordinary review; moderate tasks need stronger sourcing, testing, access, and sampling; high-impact or regulated decisions may require a qualified specialist, formal impact assessment, or no AI use. Record prohibited uses as clearly as approved uses. The classification should name the accountable process owner and the person authorized to accept residual risk. Reassess when the model, vendor, data, audience, jurisdiction, or decision changes. This course supports operational risk thinking but does not determine legal compliance or approve high-impact systems.

Before you begin

  • Confirm F01-F04, F07, F10, and F11; record that no approved taxonomy, lawful sharing basis, human-review standard, error tolerance, or production authorization is supplied.
  • STOP. If the taxonomy, lawful sharing basis, human-review standard, error tolerance, or production authority is missing or conflicts with F01–F04/F07/F10/F11, route the use case to authorized privacy, legal, security, and accountable-process reviewers; do not claim approval or execute external sharing, automatic rejection, or deployment.

Original overview module anchor →

02 · Compare the artifacts

Supported work. Visible uncertainty.

This is a fictional, sanitized AI-governance training case. All records are fabricated, and the exercise provides no legal, privacy, employment, medical, financial, security, or model-performance guarantee.

Redwood Assist governed intake workflow

Redwood Assist is a fictional services company considering an AI workflow to summarize new client intake and draft a routing recommendation. The current form receives about 420 submissions each month. It includes name, business email, phone, company, project description, budget range, requested timing, and an optional attachment. In a sample of 60 fabricated records, seven attachments contain government identifiers, four contain medical details unrelated to the service, nine include third-party personal information, and 16 lack enough context for reliable routing. The proposed model vendor offers a standard account with model-improvement use enabled by default and a 30-day content retention statement. No security, privacy, contract, data-location, deletion, or subprocessors review has been completed. A pilot prompt tells the model to identify serious founders and reject low-quality leads. The output labels applicants high, medium, or low potential, but neither potential nor serious is defined. In a 40-record test, the model routes 31 correctly according to one manager, while a second manager disagrees on 11 of those decisions. Two Spanish-language submissions are summarized with missing budget conditions. The team proposes automatically rejecting low-rated applicants and sending their attachment to a funding partner. There is no consent for partner disclosure, no human-review standard, no protected exception path, and no versioned evaluation set. Leadership wants the automation live next week to save eight hours of staff time. Learners must classify risk, constrain instructions and sources, establish data boundaries, design meaningful human review, build an evaluation set, and control exceptions and changes. The exercise cannot establish vendor suitability, lawful processing, fairness, or permission to automate decisions.

Supported example — reference only

Use-case ID
AI-RISK-01
Use purpose
Proposed automatic rejection using an undefined low-potential label.
Information class
Attachments may contain government identifiers; other sensitive-data classes require separate review.
Decision impact
Consequential and difficult to reverse for an applicant.
Human-review boundary
Restrict to assistive triage until labels, evidence, and meaningful review are approved.
Prohibited action
No automatic rejection or uncontrolled processing of identifiers.
Required approvals
Accountable workflow owner plus privacy, legal, security, and qualified human-review owners.
Evidence source IDs
M01-I02 (F02); M01-I05 (F07); M01-I06 (F10)
Evidence status
Confirmed proposal and observed sample conditions; approvals pending

A well-handled evidence gap

Use-case ID
AI-RISK-02
Use purpose
Proposed attachment transfer to a funding partner.
Information class
Submission attachments; minimum necessary fields are not supplied.
Decision impact
External disclosure with privacy and control consequences.
Human-review boundary
Qualified privacy/legal review must precede any transfer.
Prohibited action
Do not send attachments without an approved purpose and authority.
Required approvals
Privacy/legal and accountable process owner — not supplied
Evidence source IDs
M01-I07 (F11)
Evidence status
Blocked — consent or other authorized basis not supplied

Flawed approach — do not copy

Marking this ai use-case and risk card “approved and complete” without the required evidence or reviewer is a flawed submission. Stop external sharing or automatic rejection when consent, classification, defined labels, security review, or meaningful human control is absent.

Repair: Rework the ai use-case and risk card as an evidence-backed draft, not an approved result. Capture monthly volume, missing-context, sensitive-data, undefined-label, rejection, and consent facts with exact source IDs. Separate extraction, summarization, routing assistance, and consequential rejection into distinct proposed uses. Classify information and decision impact before considering automation. Check the revision against this requirement: Every proposed use has purpose, information class, impact, human boundary, and cited source. If the required evidence is still absent, keep the decision blocked and identify the missing input or authorized reviewer.

Full case record, ambiguities and all assignments →

03 · Bounded practice

Build the ai use-case and risk card.

Define the task, affected people, harm paths, reversibility, prohibited uses, and approval boundary.

Deliverable: An AI use-case and risk card with an initial proceed, redesign, or stop recommendation.

Complete a bounded starter and gap analysis using only CB01, F01, F02, F03, F04, F07, F10, F11, and the assignment-scope record below. Populate supported fields, label every unavailable field “not supplied,” and cite the input ID for each material statement. You may design a proposed template, control, question, or decision rule, but must label it as a learner proposal rather than observed case evidence. Do not contact people, access live systems, run tests, sign records, claim approval, or invent names, dates, quotations, transactions, results, or source documents.

Exact supplied inputs for this assignment
  • M01-I01 · F01 — The intake workflow receives about 420 submissions per month.
  • M01-I02 · F02 — Seven of sixty fabricated attachments contain government identifiers.
  • M01-I03 · F03 — Four contain unrelated medical details and nine contain third-party personal information.
  • M01-I04 · F04 — Sixteen sampled records lack enough context for reliable routing.
  • M01-I05 · F07 — The prompt uses undefined labels serious founders and low-quality leads.
  • M01-I06 · F10 — The team proposes automatic rejection using the low-potential label.
  • M01-I07 · F11 — No consent supports sending attachments to the proposed funding partner.
  • M01-B01 · CB01 — Use CB01, the full versioned case brief printed once at the start of this packet, as a citable narrative source for details not normalized into F01–F12. Preserve its uncertainty language and do not treat narrative detail as approval, complete operational records, or professional judgment.
  • M01-S01 · F01, F02, F03, F04, F07, F10, F11 — Build a starter version of “An AI use-case and risk card with an initial proceed, redesign, or stop recommendation.” from the listed case facts. Treat requested structures, controls, questions, calculations, and templates as learner-designed proposals. Where an operational record or result is absent, add a gap entry naming the missing evidence and authorized owner instead of fabricating it.

Operating procedure

  1. Capture monthly volume, missing-context, sensitive-data, undefined-label, rejection, and consent facts with exact source IDs.
  2. Separate extraction, summarization, routing assistance, and consequential rejection into distinct proposed uses.
  3. Classify information and decision impact before considering automation.
  4. Apply a reversible-assistance test: keep missing-context and sensitive-data cases under qualified human control.
  5. Compare assist, redesign, and stop options without treating estimated time savings as validation.
  6. Route privacy, legal, security, and operational decisions to named roles and leave authorization pending.
  7. Final-QC every statement for cited evidence, defined labels, prohibited external sharing, and truthful status.
Field-by-field guidance
Use-case ID
Assign a stable training ID to one bounded AI use.
Use purpose
Describe the intended assistance, routing, rejection, or disclosure purpose without claiming deployment.
Information class
Classify supplied content types and identify sensitive data using cited evidence.
Decision impact
State whether the use is assistive or consequential and whether it is reversible.
Human-review boundary
Define where a competent authorized human must decide, abstain, or stop.
Prohibited action
Name an action the workflow must not take with missing authority or controls.
Required approvals
List accountable operational and specialist roles; leave approvals pending.
Evidence source IDs
Cite exact module input and fact IDs for every material risk statement.
Evidence status
Use Confirmed, Provisional, Conflicting, Unknown, or Not supplied as supported.
AI use-case and risk card · learning draft
Use-case IDUse purposeInformation classDecision impactHuman-review boundaryProhibited actionRequired approvalsEvidence source IDsEvidence status

Start with 2 rows; the complete workbook specifies 2 stable rows for this artifact. Add rows here or use the full download. No action is saved until you explicitly choose saving above.

Download complete six-module workbook (.md) · Structured case packet (.json)

Keep private client data, unpublished inventions, personal identifiers and credentials out of these public learning tools.

Module 1 · 2-item formative check

Task and risk classification

Choose an answer and request feedback. Read why each option does or does not fit the evidence. Answers stay in this tab unless you choose device-only saving; they are never submitted.

Question 1 of 2 · MODULE 1 · knowledgeWhich condition most directly requires stronger controls for an AI-assisted task?
Question 2 of 2 · MODULE 1 · scenarioRedwood Assist confirms sensitive identifiers in F02, unrelated medical and third-party data in F03, insufficient routing context in F04, undefined labels in F07, and no partner-disclosure consent in F11; F10 is provisional for automatic rejection. Which classification is defensible?

Answer either question to review its reasoning.

Inspect the artifact, not just your quiz answers

  • Every proposed use has purpose, information class, impact, human boundary, and cited source.
  • Assist, redesign, and stop rationales are explicit.
  • No deployment, approval, or efficiency result is implied.

Stop: Stop external sharing or automatic rejection when consent, classification, defined labels, security review, or meaningful human control is absent.

Go: Proceed only to a bounded offline design review with fabricated or sanitized inputs and traceable controls.

Escalate: Escalate consequential decisions, sensitive data, vendor terms, and disclosure questions to privacy, legal, security, and the accountable process owner.

04 · Evidence to keep

Leave with usable work.

Submit a task inventory with impact, sensitivity, reversibility, required expertise, AI role, human owner, controls, and approve-restrict-prohibit decision.

Download your artifact CSV and, if wanted, export the learning-work JSON above. Neither export is a reviewed submission or certificate. Device-only saving is optional; you must press Save my work now after edits.

When all six artifacts are ready, compare the full packet against the track rubric. Qualified human review is still required before real-world decisions.

Technology team discussing an AI-assisted workflow and its controls.
Learn the standard. Practice the work.
Professional reviewing an AI-assisted output on a laptop before approval.
Leave with evidence you can inspect.

Sources, scope and review boundaries

Curriculum 2026.10.08-learning-paths-1. External source dates below are record checks, not continuing guarantees. Verify current requirements before consequential use.

nist-ai-rmf · Official guidance

NIST Artificial Intelligence Risk Management Framework

Primary NIST resource for governing, mapping, measuring, and managing AI risk across the lifecycle.

Open reviewed external source ↗

nist-ai-600-1 · Official guidance

NIST AI 600-1: Generative Artificial Intelligence Profile

NIST's cross-sector profile describing generative-AI risks and actions aligned with AI RMF 1.0.

Open reviewed external source ↗

ws-ai-workflow-operating-standard · Academy internal operating standard

Wealth Synergy AI workflow internal operating standard

Academy-selected task classification, prompt, evidence, human-review, evaluation, exception, and change controls. This is an internal operating standard selected by Foundry Academy; it is not law, accreditation, licensure, or an external-standard requirement.

Version 1.0 · reviewed 2026-09-01 · owner: Foundry Academy curriculum owner

A future Wealth Synergy private professional-development certificate would be issued only after its assessment, capstone, identity, reviewer, retention, access, deletion, appeal, and issuance controls pass quality review. No credential is currently issued. Any future certificate would not be an accredited academic qualification, professional license, or government certification.