He was stopped after leaving an electronics retailer.
An officer demanded proof that the computer belonged to him.
The accountant showed the receipt.
The officer accused him of being argumentative.
He was detained while the transaction was verified.
Again, no crime was established.
Again, the system recorded a suspicious-person contact.
A third case involved a Black mother shopping with her children.
She was accused of using a fraudulent gift card.
The card was valid.
But the confrontation frightened her children so badly that she stopped visiting the shopping center.
When she complained, the retail-security contractor responded that its employees had acted out of an abundance of caution.
The phrase appeared repeatedly.
Abundance of caution.
The judge stared at those words.
She thought about the young cashier who had risked her job to provide the duplicate receipt.
Then she thought about the other customers.
How many had been falsely accused?
How many had simply gone home ashamed?
How many had never filed complaints because they believed nobody would listen?
The judge knew the answer could not come from anger alone.
It required reliable records.
Independent review.
And the willingness of institutions to acknowledge when their procedures were causing harm.
The investigation into the retail-security system expanded.
Auditors examined how alerts were classified.
Independent experts reviewed the risk-scoring methods.
Investigators compared flagged transactions with actual evidence of theft or fraud.
The results were deeply concerning.
Some alerts were based on legitimate information.
But others relied on subjective judgments unrelated to specific criminal behavior.
The reporting categories allowed employees to mark an intervention as successful even when no theft or fraud was established.
This created a distorted picture.
The company could claim that its system prevented crime while counting many encounters involving innocent customers.
Worse, the records showed significant demographic disparities requiring explanation and corrective action.
The system was not merely suffering from a few clerical errors.
Its incentives encouraged employees to generate suspicious contacts.
The more contacts recorded, the more active and valuable the service appeared.
And when those contacts involved people whom staff already viewed with suspicion, the system reinforced the very assumptions it should have been designed to test.
The compliance officer's testimony became particularly important.
She explained the internal reporting structure.
The performance metrics.
The warnings she'd raised.
And the executive pressure to count unsubstantiated contacts as successful interventions.
Her records showed that concerns about discriminatory impacts had been discussed months before the judge's encounter.