Necessary.
That was how power always introduced itself.
The first public criticism of the Leadership Ethics Review Board came from a colonel Ava respected.
Colonel Patrick Doyle commanded a brigade in Georgia.
He had no ethics findings.
No retaliation complaints.
No disciplinary history.
His units performed well.
His command climate surveys were above average.
And he hated Mirror.
Not quietly.
He published an article in a professional military journal titled:
WHEN OVERSIGHT BEGINS TO COMMAND
Ava read it at 0530 before physical training.
Then again after breakfast.
Doyle’s argument was simple.
Oversight systems begin as protection.
Then they accumulate authority.
They gain access.
They create reporting requirements.
They influence promotions.
They shape what commanders believe they are allowed to say.
Eventually leaders stop using judgment because they are afraid someone outside the chain will reinterpret it.
Doyle wrote:
A commander who must constantly imagine how an external reviewer might interpret every difficult decision is no longer commanding freely. He is managing perception for auditors.
Ava hated that sentence.
Mostly because part of it was true.
Miriam Shaw called at 0705.
“You read it?”
“Yes.”
“And?”
“He is wrong.”
“Completely?”
Ava paused.
“No.”
“Good.”
“I still dislike him.”
“You have never met him.”
“I dislike the article.”
“Better.”
By 0900, the article had circulated across senior commands.
Some leaders agreed openly.
Others privately.
One general called Mirror “a parallel command structure.”
Another called it necessary protection after repeated retaliation failures.
A retired command sergeant major wrote that junior soldiers needed independent channels because chains of command naturally protect themselves.
The debate became larger than any single case.
General Creel, no longer board chair but still influential, called Ava.
“Do not respond publicly.”
“I wasn't planning to.”
“You were thinking about it.”
“Yes.”
“Don't.”
“Why?”
“Because you are too close.”
Fair.
Miriam, acting as civilian co-chair, convened internal review.
Not about Doyle.
About Mirror.
Question:
Had the board accumulated influence beyond its original mandate?
Ava disliked that question even more.
Good.
They mapped authorities.
Complaint intake.
Independent investigation.
Access to protected records under specific conditions.
Corrective recommendations.
Mirror triggers.
Training guidance.
Policy consultation.
Pattern analytics.
Over seven years, the board had become involved in more areas than originally designed.
Not because someone planned takeover.
Each expansion followed a real problem.
A data-access failure created cybersecurity review authority.
A retaliation case created stronger complaint protections.
A remediation dispute created restoration procedures.
A commander-climate case created follow-up assessment.
Reasonable.
Together, large.
Miriam said:
“This looks familiar.”
Ava nodded.
“Every layer was justified.”
“Exactly.”
Major Grant looked toward the authority map.
“We may have built bureaucracy.”
Ava remembered Maya Patterson from another kind of organization she had once studied in a civilian case packet.
Same pattern.
Good safeguards accumulate until nobody asks whether each still earns its cost.
The board conducted anonymous commander interviews.
Results uncomfortable.
Some leaders valued Mirror.
Others feared it.
Several said they no longer gave candid developmental feedback about “aggressiveness,” “resilience,” or “executive presence” because those words had become associated with bias review.
Ava frowned.
“That is not what guidance says.”
Miriam replied:
“Does not matter if leaders believe it.”
Another pattern:
Some commanders consulted ethics staff before ordinary personnel decisions simply to create protective record.
Not because they needed advice.
Because they wanted future defense.
Defensive administration.
Bad.
One colonel said:
“I don't need permission, but I want proof someone else agreed.”
There.
Oversight creating dependency.
The board had wanted leaders to think more carefully.
Some had learned to outsource judgment.
Ava hated that.
They invited Patrick Doyle.
He arrived skeptical.
No apology.
Good.
Miriam opened.
“You believe the board has too much influence.”
“Yes.”
“Specific.”
Doyle brought examples.
One company commander delayed corrective counseling three weeks while waiting for ethics consultation.
Not required.
Still.
Another unit added external review to a routine reassignment because leaders feared retaliation complaint.
Again, not required.
Another commander avoided removing a toxic but popular first sergeant because he worried action would appear retaliatory after first sergeant had filed unrelated safety complaint.
Real dilemma.
The commander eventually acted.
Too slowly.
Doyle said:
“Your system is teaching commanders that any person who complains becomes untouchable.”
Ava replied:
“That is not policy.”
“I know.”
“Then that's a training problem.”
“No.”
Doyle looked directly at her.
“It is a power problem.”
Silence.
Ava leaned back.
“Explain.”
“People respond to institutions based on perceived consequence.”
“You may tell commanders they retain discretion.”
“But they know your findings can reach promotion boards.”
“Can trigger external review.”
“Can become permanent record.”
“So they protect themselves.”
Ava disliked him less now.
Because he was making argument she could not dismiss.
“What do you recommend?”
“Reduce advisory involvement.”
“Mirror investigates serious concerns.”
“It does not become pre-clearance service.”
“Limit pattern analytics.”
“Why?”
“Because when everything becomes data, normal friction starts looking systemic.”
That one hit.
Ava had lived that danger personally.
“Anything else?”
“Yes.”
“Put expiration dates on new safeguards.”
Miriam looked interested.
Doyle continued.
“Every new requirement sunsets unless evidence shows it still solves more than it costs.”
Ava almost smiled.
“That sounds annoyingly reasonable.”
Doyle did not smile.
“I'll survive.”
The board reviewed.
Some authority stayed.
Independent complaints.
Retaliation review.
Unauthorized access triggers.
Evaluator accountability.
Mandatory intervention in active harm.
Core.
Other functions reduced.
Routine advisory consultation moved back to normal legal and command channels.
Mirror could not “pre-approve” ordinary personnel decisions.
Pattern analytics thresholds raised.
De-identified screening required before investigators could see names.
New policy safeguards received review dates.
The board also adopted a rule:
Oversight shall not substitute for command judgment where no specific ethical, legal, or retaliation risk is identified.
Ava supported.
Publicly, Mirror issued its own scope reduction.
Some critics called victory.
Fine.
Some supporters called retreat.
Also fine.
Doyle wrote follow-up article.
Less dramatic.
He acknowledged changes.
Still skeptical.
Good.
Ava later met him at conference.
“You still think Mirror should exist?”
He considered.
“Yes.”
“That surprise you?”
“A little.”
“Why?”
“Because organizations need someone who can look outside chain.”
Ava nodded.
“And commanders need space to command.”
“Yes.”
“Both.”
“Yes.”
Doyle smiled.
“Reality is inconvenient.”
Ava laughed.
She had heard that before in different forms.
The confrontation mattered because Mirror learned a system can become too powerful without anyone abusing it intentionally.