03 / ASSURANCE ENGINEERING

Control Intent.
Make it true.

Is the control achieving its purpose? Control Intent connects intent to observable evidence, so you can distinguish a working control from a convincing record.

Work through an example

A practical method for connecting control purpose, behavior, and evidence.

  1. 1. Risk
  2. 2. Intent
  3. 3. Behavior
  4. 4. Evidence
  5. 5. Gap
  6. 6. Correction
  7. 7. Validation
The assurance loop intent, operation, learning.
THE DISTINCTION

Proof and performance
are different questions.

A record can document a review while leaving the underlying access unchanged. A control can also work while its evidence is inadequate.

Identify which problem you have before asking for another submission. Improve the behavior when it is wrong; improve observability when its operation cannot be established.

I don't want better evidence of a broken system. I want the intended outcome to become true by construction.

Control Intent begins with the outcome, works back to the mechanism, and forward to evidence that can demonstrate it.

THE METHOD

Seven questions.
One correction loop.

  1. Risk

    What harmful state must be prevented?

  2. Intent

    What needs to remain true?

  3. Behavior

    What does the system actually do?

  4. Evidence

    What can the record establish?

  5. Gap

    Where does behavior or visibility fall short?

  6. Correction

    Which mechanism or handoff needs to change?

  7. Validation

    How will we verify that the correction holds?

HYPOTHETICAL WORKED EXAMPLE

Access reviews should
change access.

Suppose the intended outcome is that unauthorized access does not persist. A signed quarterly review shows an activity took place. It does not, by itself, establish that every removal was completed.

Illustrative reasoning, not an employer incident or a completed control assessment.

Inspect the prepared evidence investigation
Intended outcome
Access no longer authorized is removed.
Observed behavior
A reviewer marks accounts for removal.
Available evidence
A signed review and the list of decisions.
What it establishes
Who reviewed the list and which decisions were recorded.
What remains unknown
Whether removals occurred across every in-scope system and within the required time.
Next validation
Trace decisions to removal records, check scope and timing, then investigate exceptions.
FROM QUESTION TO ACTION

Fix the mechanism.
Then verify the result.

If work stalls at a handoff, clarify ownership. If behavior is wrong, change the system. If the record cannot show the outcome, collect evidence closer to the source.

The aim is a repeatable operating condition, with a clear way to know whether it holds.

Try it on the public example.

Assurance Navigator includes a prepared access-review investigation. Compare the artifacts, missing requirements, and analyst judgment.

Synthetic, prepared scenario. It illustrates evidence reasoning; Control Intent is a method, not a complete control-testing product.

Inspect the investigation
BRING A CONTROL QUESTION

What should be true
in your system?

We can discuss the intended behavior, what the evidence supports, and where to investigate next.

Discuss a control question