30. Camille
Camille
The conference room is colder than it should be. Not physically. The temperature is maintained at a precise, comfortable degree, and the lighting is soft enough to reduce strain during extended meetings. Everything has been optimized.
Everything except the people inside it.
I take my seat without waiting to be directed. That, at least, remains within my control.
Devin is already there, posture straight and hands folded neatly on the table. He looks composed, but the set of his mouth suggests he has spent the last several hours explaining Tokyo to sponsors who do not appreciate supernatural nuance.
Senior Councilman Eric Cohen stands near the far wall rather than sitting. He prefers movement when managing variables, though he disguises the habit as impatience.
Blaire arrives last. She closes the door with deliberate care, as though sealing the room might keep whatever happens inside it contained.
No one speaks immediately.
They are waiting for my assessment.
Good.
“I found several inconsistencies in the Tokyo data,” I say. “Before I finalize the incident report, I need to determine which of them were expected.”
Cohen’s attention sharpens. Devin remains still.
Blaire watches me quietly.
The incident report is an excuse. I need to determine which of them recognizes the architecture before I reveal anything further.
“What inconsistencies?” Devin asks.
I connect my tablet to the table display and project a simplified version of the damaged contract architecture. Not everything. I have no intention of revealing how deeply I accessed the system or how much I have already connected.
Bane’s preliminary report already confirms that I accessed the restricted architecture. Concealing the clause entirely would achieve nothing. Better to let them believe I found one malfunction than the structure beneath it.
“The emotional output exceeded every projected threshold during the second chorus,” I say. “That could be attributed to the Mouthpiece interference. This cannot.”
I isolate a timestamped capture from the Tokyo concert.
EMOTIONAL YIELD OPTIMIZATION — ACTIVE
Devin’s eyes flick towards it before moving away again.
Too quickly.
Interesting.
“This remained active after the performance exceeded containment limits,” I continue. “It also remained active during Jules’s possession. Is that standard behavior under the legacy framework?”
Cohen steps closer to the table. “The clause is part of an older engagement model.”
“That does not answer the question.”
His expression barely changes. “The system is designed to maintain continuity during temporary instability. Otherwise, every technical interruption or minor breach would collapse the containment field.”
A technically reasonable explanation.
It does not account for the extraction pattern.
“And the emotional amplification?” I ask.
“Higher audience engagement creates a stronger resonance draw.”
“I understand the draw. I’m asking why the optimization process continued after that draw became unsafe.”
Cohen glances at Devin.
A small movement. Easily missed.
“We had models suggesting the system would regulate itself,” Devin says.
There it is.
Not an admission. Not quite.
Enough.
“What did those models define as successful regulation?”
“Containment of the gathered demonic energy.”
“And performer stability?”
“Part of the same calculation.”
“Weighted equally?”
Neither of them answers immediately.
I allow the silence to remain.
Cohen moves closer to the display. “Midnight Halo’s resonance profile is unusually high. The scale of the Tokyo engagement event created conditions we have not observed before.”
“That is not entirely accurate.”
His gaze lifts to mine.
I keep my expression neutral.
“Barcelona produced an abnormal spike,” I say. “London produced a possession attempt through infrastructure. Seattle showed altered musical layers responding to crowd output. Tokyo was larger, but it was not without precedent.”
“Those incidents were reviewed individually,” Devin says.
“Yes.”
I leave the word there.
Individually.
That is how the system protects itself. Every failure assigned a separate explanation. Every pattern divided until no one is responsible for recognizing the whole.
I change the display.
RESIDUAL RESONANCE RETENTION — 18.4%
The number is from Tokyo, recorded moments before I closed the primary pathway. I have adjusted the label to conceal the system’s original terminology.
“Is this expected?” I ask.
Cohen studies the figure. “Some resonance remains in the ward network after high-output performances.”
“For how long?”
“It disperses.”
“That was not the question.”
His jaw tightens slightly. “Duration varies.”
“Based on what?”
“Event scale. Audience response. Environmental conditions.”
“And where does it disperse?”
This time, he does not answer.
Devin leans forward. “What exactly are you suggesting?”
I keep my tone even. “I am attempting to determine whether Tokyo’s residual output indicates system damage or established function.”
It is the truth, but not all of it.
Cohen looks back at the number. “Retention was historically considered a containment by-product.”
“Historically?”
“The models are old. The framework was never fully replaced.”
“Why?” I press.
“Because it works.”
The answer comes too quickly.
Blaire shifts beside me.
Only slightly, but I notice.
I look at Cohen. “Tokyo does not appear to support that conclusion.”
“The arena was evacuated without mass casualties,” Devin says. “The Mouthpieces were neutralized. The official narrative is holding.”
Survival presented as proof of success.
Predictable.
“And Midnight Halo?” Blaire asks.
It is the first time she has spoken.
Devin turns towards her. “All four members survived.”
“That wasn’t what I asked.”
His expression cools.
Blaire does not look away.
I intervene before the exchange becomes useful to no one. “The group survived because the primary pathway was terminated before the possession stabilized.”
Cohen’s attention returns to me. “An emergency override that caused significant structural damage.”
“Yes.”
“You accessed Council-restricted architecture.”
“The standard controls were unresponsive.”
“And if terminating that pathway had destabilized the bond?”
“It did not.”
“That was not guaranteed.”
“Neither was Jules’s survival if I left it open.”
The words come closer to accusation than intended.
I adjust.
“My concern,” I continue, “is whether the system’s safeguards responded as designed. If they did, then the design requires review. If they did not, we need to identify the failure before Midnight Halo performs again.”
Cohen studies me.
He is trying to determine whether I have reached the truth or merely brushed against it.
I let him wonder.
“The Mouthpieces exploited multiple access points,” I say. “Production staff. Broadcast systems. Audience synchronization. Has that pattern appeared in previous incident reports?”
“Not at that scale,” Cohen says. “There have been isolated events.”
Blaire’s gaze shifts towards him. “And managers were not fully informed?”
“Threat assessments are distributed according to operational need.”
“I manage the performers those threats are targeting. I would consider that operational need, Mr. Cohen,” Blaire says sharply.
“You receive the information required to keep them safe.”
Blaire’s expression does not change, but anger enters the room. She knows the language. She has likely heard variations of it for years.
“Did any previous event involve a larger presence operating through the conduits?” I ask.
Blaire looks at me, not with recognition but with alarm.
Good.
She does not know.
Cohen answers carefully. “Certain reports reference parent entities.”
“Associated with previous Eon groups?”
His gaze hardens.
There is the boundary.
“We are not operating with complete information,” he says.
“Does the Council have complete information?”
“I cannot speak for the full Council.”
Cannot.
Not will not.
Useful.
I close that portion of the display before he can decide I have asked too much.
“I need the operational summaries from every related incident and continued access to Midnight Halo’s contract telemetry.”
“That exceeds your clearance,” Devin says. “Assuming your clearance stays the same.”
“The residual network is still active and rewriting itself around the pathway I terminated. Remove my access now, and you interrupt the only analysis capable of tracking it.”
Cohen looks towards Devin. “She’s right.”
“Mr. Bane’s preliminary report also recommends continuity,” I add. “He believes changing personnel while the residual network is recalibrating would create unnecessary risk.”
Bane’s wording was more clinical.
The meaning was the same.
Devin’s mouth tightens. “Your attachment to Midnight Halo will remain in place pending review.”
Not approval.
Not trust.
Enough.
“I also recommend temporary suspension of any Midnight Halo performances. The Tokyo broadcast is currently being explained as a catastrophic technical malfunction. Repeating the conditions responsible would be difficult to defend to sponsors.”
It’s language that Devin will understand.
“Midnight Halo’s purpose doesn’t allow for indefinite suspension,” he says.
“I am not recommending an indefinite suspension.”
“Any meaningful suspension will reduce engagement,” Devin says.
“It will also prevent a second Tokyo.”
Cohen studies the closed display and changes the subject. “A temporary suspension is reasonable while the Tokyo data is reviewed.”
Devin does not agree, but sponsors are already questioning why Jules performed while injured.
“Submit the protocols by morning,” he says.
For now, the girls have space.
“That is sufficient,” I say.
It is not, but I stand as though it is.
The meeting is over.
No one stops me. No one questions my continued role. I leave as the person who contained Tokyo, preserved the public narrative, and identified a technical vulnerability before it could recur.
Useful.
Concerned.
Cooperative enough to retain.
They do not believe I am harmless. They believe I am still containable.