Page 191 - CyberSecurityEssentialsEnglish

The Practical Lab: Run a Real Task Through 4D
Choose a useful task of moderate complexity, such as summarizing an internal report,
drafting a plan, or turning meeting notes into tasks. Do not begin with a highly sensitive task.

The goal is to build a habit you can repeat, not to test the outer limits of the tool.

Stage one: write the use definition card and set the level of impact. Create a safe copy of the
data. Explain what you removed or replaced and why. If you cannot minimize the data, use
public or synthetic material for the lab.

Stage two: set the ladder of delegation. Let the system produce a draft or an extraction, and
do not grant it external execution. Write down the approval point: who reviews, and which
items must be checked?

Stage three: extract five claims from the output. Verify at least two sources, and use SIFT for
an unfamiliar source. Test a borderline case or an example that differs in language or

phrasing. Record what was confirmed, what changed, and what was removed.

Stage four: complete the short 4D record.
Write a decision in four sentences: will the output be used? In what scope? What are its

limits? And who approves it? Then estimate the time the tool saved after adding review and
correction time.

TABLE 4-20: LAB OUTPUTS AND THE CRITERION OF SUCCESS

Output Criterion of success

It states the purpose, the impact, the data, and the
The use definition card
stopping condition.

A safe copy of the material It contains no unnecessary data and no direct identifiers.

It defines the role of the tool, the role of the human, and
The delegation plan
the approval point.

The claim and evidence
It ties every claim to a source and a verification status.
table

It records the tool, the sources, the changes, and the
The 4D record
decision.

The risk-adjusted
It computes net time and states the residual risks.
productivity estimate

190