The Black Box Society: The Secret Algorithms That Control Money and Information: Summary & Key Insights

When I began studying information law, algorithms were still understood as tools—mathematical procedures designed to help humans solve defined problems. Today, they are becoming governors in their own right, executing judgments about who deserves credit, what deserves visibility, and which threats merit attention. This transformation has been powered by the combination of vast data collection, machine learning, and proprietary systems owned by a handful of private entities.

Corporations now promise objectivity and efficiency through automation, but the very opacity of these processes conceals how human bias, market incentives, and structural inequality become encoded. Think of predictive policing systems that target marginalized communities because historical data reflects over-policing in the past. Consider targeted advertising engines that deliver job postings for high-paying positions predominantly to men. These are not random accidents but logical outcomes of algorithms trained on biased data. And because the criteria are secret, those affected cannot contest their exclusion.

Governments too have adopted algorithmic decision-making in welfare distribution and national security screening, invoking 'data-driven governance' to justify efficiency gains. Yet, the more consequential their deployment, the less transparent they become, hidden behind both national security and corporate confidentiality. What emerges is a new form of power—algorithmic authority—which appears neutral but often replicates and deepens existing hierarchies.

Finance is perhaps the quintessential black box industry. Algorithmic trading now dominates markets with unimaginable speed, executing thousands of trades in microseconds, yet few outside the trading firms understand how decisions are made. When the 2010 'Flash Crash' briefly erased nearly a trillion dollars in market value, regulators struggled to trace accountability—an episode that perfectly illustrates my thesis: complexity can become a shield for irresponsibility.

At a personal level, algorithms also determine our access to credit and insurance. Credit scoring systems, once based on simple metrics, now draw on behavioral and even social network data, generating risk profiles few consumers can interpret or challenge. Entire financial lives are thus governed by invisible formulas ranking our 'worthiness.'

Such opacity has profound consequences for economic justice. Just as discriminatory redlining once excluded minorities from mortgages, algorithmic models can reproduce bias under a veneer of neutrality. Yet, because they are protected as trade secrets, neither regulators nor consumers can peer inside. The lesson is clear: when financial logic and technical secrecy merge, democratic oversight evaporates, and the market becomes not a forum of fairness, but one of silent exclusion.