He returned to the study instead of bed. The desk was nearly clear except for his laptop, the photograph of his father, and a stack of documents from the closing. He opened an encrypted folder labeled Hullbrook Dynamics—Majority Acquisition. For several minutes he read the summary page without touching the keyboard.
Ellison Holdings owned fifty-one percent of Hullbrook Dynamics. The transaction price was $47.3 million, with the purchase completed on October fifteenth after months of confidential negotiations. Most of the shareholders and several directors knew the deal existed, but the identity of the principal behind the holding company had not been disclosed widely. The Hullbrook family retained a significant minority stake and several seats on the board. Under the restructuring agreement, Marcus had broad inspection rights, the authority to propose changes in executive oversight, and the voting support needed to replace the company’s failing leadership if his concerns were confirmed.
The company had once been an impressive engineering business. Its earliest products helped cities maintain water systems and keep municipal records safe during disasters. Under Arthur Hullbrook’s direction, however, the firm had chased a more profitable market in predictive public safety technology. Its flagship platform, SafeWatch, combined neighborhood reports, camera-generated alerts, contractor access lists, address data, and other information into scores that police dispatchers could use to prioritize responses. In sales presentations it sounded efficient and modern. In complaint files, the picture was much less reassuring.
Marcus had come across the company while reviewing municipal software contracts for another investment. At first he noticed unusually high renewal costs and clauses that shielded the company from responsibility when its alerts proved inaccurate. Then an analyst found a pattern in complaints from affluent neighborhoods. Workers who entered gated communities regularly were reported as suspicious even when they had uniforms or scheduled appointments. Black and Hispanic visitors appeared disproportionately among the people assigned high concern scores. Several residents reported being questioned near their own homes after the system failed to recognize their new addresses.
None of that automatically proved that every employee had intended discrimination. Marcus knew complex systems could cause harm through careless design, bad training data, or incentives that rewarded the wrong results. But he also knew executives sometimes preferred to call those harms accidents because an accident sounded easier to forgive than a profitable pattern. He had continued digging. The company’s collapsing finances gave him an opening to purchase a controlling interest and demand the records necessary to understand what had happened.
His advisers had questioned the risk. Why spend tens of millions on a troubled company carrying lawsuits and public distrust? The answer was partly financial; Hullbrook retained valuable engineers, infrastructure contracts, and technology that could be rebuilt. It was also personal. Marcus had spent his career learning what happened when a technical decision in a distant office changed someone’s day on a real street. He believed such systems could be made accountable, but only if the people who owned them stopped pretending that code erased responsibility.