Level 1 · Guide 01
Why can Claude be confidently wrong?
Claude can produce incorrect facts or unsupported citations while trying to be helpful. For consequential decisions, never treat one answer as the sole source of truth; inspect the original source, date, numbers, and context.
Product features may change. Check the last verified date and official sources.
Core concepts to know first
Focus on the decisions these terms support in real work rather than memorizing them.
Core concepts
1. Fluency is not evidence
Fluent language is not evidence. Separate every answer into claims, cited support, assumptions, and unknowns, then match the strength of verification to the cost of being wrong. In this guide, the first observable move is to split an answer into checkable claims and assumptions.
2. Three information layers
Open each cited source and compare the exact passage. Preserve the approved input and the evidence behind the result so a reviewer can distinguish what the source says from what Claude inferred. This directly controls the risk that a real-looking citation may not support the sentence.
3. Unknown can be a good result
Record a use, revise, or reject decision for every important claim. Record the decision and the remaining uncertainty instead of hiding it in polished prose. The review must explicitly test whether repeated answers can repeat the same error.
4. Verification should match risk
Treat the possibility that a confident tone can hide missing or outdated evidence as a required test case. The intended result is a one-page claim and evidence verification card, not an unreviewed answer that merely looks complete.
Synthetic work scenario
How this applies at work
This scenario was written for learning and is not a real customer case.
A fictional Korean planning team asks Claude for a market figure before an executive meeting. The team labels each number, opens the cited source, checks the publication date, and marks one unsupported claim as unknown instead of polishing it.
The scenario is newly written for this guide and is neither a customer case nor a performance claim. Its specific deliverable is a one-page claim and evidence verification card. A reviewer can reproduce the work from the synthetic inputs without access to customer, employee, health, contract, or confidential company data.
Try it yourself
Choose one small task and follow the steps. Confirm organizational policy and data boundaries before using sensitive materials.
Step 1. Split an answer into checkable claims and assumptions
Split an answer into checkable claims and assumptions. Use only approved synthetic material and record both the evidence and any remaining uncertainty.
Verify: Confirm that another reviewer can reproduce the input, result, evidence, and stop point.
Step 2. Open each cited source and compare the exact passage
Open each cited source and compare the exact passage. Use only approved synthetic material and record both the evidence and any remaining uncertainty.
Verify: Confirm that another reviewer can reproduce the input, result, evidence, and stop point.
Step 3. Record a use, revise, or reject decision for every important claim
Record a use, revise, or reject decision for every important claim. Use only approved synthetic material and record both the evidence and any remaining uncertainty.
Verify: Confirm that another reviewer can reproduce the input, result, evidence, and stop point.
Completion checklist
Check only what you verified yourself. Every item must be checked before saving completion.
Some items are still unchecked. Review the result again.
What could go wrong?
Plausible language does not guarantee accuracy. Compare the result with originals, calculations, permissions, and current information.
- A real-looking citation may not support the sentence
- Repeated answers can repeat the same error
- A confident tone can hide missing or outdated evidence
Boundaries that require human review
Academy practice stops at draft, preview, or approval pending. Actions with real impact require separate owner approval outside Academy.
What AI can do
- Step 1. Split an answer into checkable claims and assumptions
- Step 2. Open each cited source and compare the exact passage
- Step 3. Record a use, revise, or reject decision for every important claim
What a person must approve
- Applicable law, contract, organizational security policy, and explicit approval boundaries take priority. Stop and ask the responsible owner when they conflict.
- Academy exercises never send, publish, purchase, delete, execute contracts, or change permissions. A responsible person performs any real action through a separate process.
This guide is educational and does not replace legal, security, or privacy judgment for your organization.
Questions about this guide
- When is this guide complete?
- It is complete when the stated outcome is ready and another reviewer can retrace the inputs, evidence, boundaries, and decision. The target outcome is “A one-page claim and evidence verification card.”
- May I practice with real company data?
- No. Use synthetic material in the Academy. Real data requires a separate review of policy, legal basis, contracts, minimization, retention, deletion, and the approved environment.
- What if the product screen differs from this guide?
- Product behavior can change. Check the verification date and official sources, then retest the current account and plan with a small, low-risk example.
Official sources and further reading
Recheck the current product and policy status in these primary sources.
Authorship and review
Save progress
Completion and checklist items are saved only in this browser. They do not sync to other devices or browsers.
Academy does not collect or store work materials, prompts, or outputs. It runs no separate analytics scripts beyond basic server access logs.