She knew him too well.
“That too.”
He finished soup.
The following morning, the young teller woke to 114 unread messages.
Some from coworkers.
Some journalists who had somehow found her social media.
Some strangers.
One message read:
YOU’RE THE BRAVE TELLER WHO STOOD UP TO HIM. INTERVIEW REQUEST.
She deleted.
Another:
We’ll pay $2,000 for exclusive.
Deleted.
Her mother called.
“You’re not talking to reporters.”
“No.”
“Good.”
Then her mother said:
“You did right.”
The young teller started crying.
She had not cried at work.
Not when the manager threatened her.
Not when corporate arrived.
Not when investigators interviewed.
Now, hearing her mother, everything broke.
“I reported months ago.”
“I know.”
“They did nothing.”
“I know.”
“I started thinking maybe I was wrong.”
Her mother was quiet.
Then:
“You saw what you saw.”
The young teller wiped tears.
“That sounds simple now.”
“It wasn’t.”
She took paid administrative leave.
Not because suspected.
Because investigators needed preserve workplace relationships while interviews occurred.
She hated it.
It felt like punishment.
The old CEO called personally.
“I don’t want leave.”
“You’re not being disciplined.”
“I know.”
“Then why?”
“Because I want to work.”
He understood.
Work can be stability.
Especially when chaos makes everyone else treat you like witness instead of person.
He offered temporary assignment at another branch.
She accepted.
“But no promotion.”
He smiled.
“I didn’t offer.”
“I’m serious.”
“So am I.”
“I don’t want anybody saying I got rewarded for going viral.”
He looked at her through video call.
“That is smart.”
Then:
“When promotion comes, it should be because you earn.”
She nodded.
“Thank you.”
“And if anybody implies you owe bank public gratitude, call me.”
She almost laughed.
“Direct?”
“Direct.”
“You’ll regret.”
“Probably.”
The external auditors discovered forty-three questionable adjustments in the first ten days.
Not all theft.
Eleven were legitimate corrections poorly documented.
Seven were teller errors.
Four customers had miscounted.
Three involved duplicate receipts.
But eighteen required deeper investigation.
Nine showed strong evidence of intentional under-crediting.
Six customers were over seventy.
Three were between fifty-five and sixty-nine.
Race varied.
Seven Black.
One Latino.
One white.
Two had limited English proficiency.
Five lived alone.
Almost all maintained relatively low balances.
The pattern was not “steal from Black customers.”
It was “steal where resistance seemed cheapest.”
Race influenced that calculation.
So did age.
Language.
Balance size.