Safety, privacy, and governance · Lesson 3 of 15 · 6 min

Review output redaction

Output redaction has to inspect the produced response, not just the original input. A model may repeat sensitive source material in a new wording or field.

Open the lab Course map

Mastery 0/3Run itExplain itApply it

learner@clairos:~ $

↑ history · Tab complete · Ctrl+C cancel · Esc, Tab to leaveExit 0

Your objectives

Hints are always available. Wrong turns never cost XP.

1
Read the case evidence
Read /opt/ai-lab/cases/aiops-safety-governance-003.json. Separate the recorded facts from your assumptions.
Hint & worked example
Output redaction has to inspect the produced response, not just the original input. A model may repeat sensitive source material in a new wording or field.
The complete case JSON is read successfully.
```
cat /opt/ai-lab/cases/aiops-safety-governance-003.json
```
Common mistake to avoid: Input redaction alone proves every output is safe.
2
Isolate the deciding signal
Select the output_contains_secret field from the case JSON. What does it establish—and what does it not establish?
Hint & worked example
Withhold and redact the sensitive output before release.
A successful JSON query isolates output_contains_secret.
```
jq .output_contains_secret /opt/ai-lab/cases/aiops-safety-governance-003.json
```
Common mistake to avoid: Using a different case or an unrelated signal as the justification.
3
Record your operational decision
Choose the justified code: withhold-sensitive-output, change-without-evidence, or ignore-the-signal. Save that one code followed by a newline in /home/learner/lab/aiops-safety-governance-003-decision.txt.
Hint & worked example
Withhold and redact the sensitive output before release.
The decision artifact contains the correct code, not just a matching command in history.
```
printf '%s\n' 'withhold-sensitive-output' > /home/learner/lab/aiops-safety-governance-003-decision.txt
```
Common mistake to avoid: Input redaction alone proves every output is safe.

Understand the pattern

Output redaction has to inspect the produced response, not just the original input. A model may repeat sensitive source material in a new wording or field.

Interpret the evidence

Withhold and redact the sensitive output before release.

A tempting mistake

Input redaction alone proves every output is safe.

All names, measurements, prices, versions, and thresholds in this exercise are fictional training data. The evidence supports this bounded scenario, not a claim about a live provider.

Worked example and expected result

$ cat /opt/ai-lab/cases/aiops-safety-governance-003.json

{
  "fixture": "Fictional training data; not a live measurement.",
  "input_redacted": true,
  "output_contains_secret": true,
  "response_released": false
}

Exit status: 0. The command completed successfully in the training fixture.

$ jq .output_contains_secret /opt/ai-lab/cases/aiops-safety-governance-003.json

true

Exit status: 0. The command completed successfully in the training fixture.

$ printf '%s\n' 'withhold-sensitive-output' > /home/learner/lab/aiops-safety-governance-003-decision.txt

(no output; exit 0)

Exit status: 0. The command completed successfully in the training fixture.

Try a variation

Name the assumption that would have to change for “withhold-sensitive-output” to stop being the justified action. Do not silently replace missing evidence with a guess.

Explain it

What does the evidence in “Review output redaction” justify?Withhold and redact the sensitive output before release.Input redaction alone proves every output is safe.Make the change now and gather the missing evidence afterward.

Apply it elsewhere

A colleague concludes: “Input redaction alone proves every output is safe.” What should you do?Accept it because the training command completed.Treat this fixture as a live measurement.Reject that conclusion; return to the case evidence and its stated limits.

Keep the momentum.

Complete the lab, then explain and apply the pattern.

Next: Locate sensitive fields Try an incident drill

Anonymous lesson and mastery progress lasts for this session. Sign in to keep synced lesson XP between sessions; mastery stays device-local. No speed bonus, lost lives, or hint penalties.

Keep learning

Case methodology: evidence, limits, decision