Extract of the Conclusion

Conclusion
The growth of algorithmic decision-making presents immense opportunity and risk to society. Algorithms could promote economic justice by helping distribute opportunities more broadly, resources more efficiently, and benefits more effectively. But this article documents the perilous potential for algorithms to amplify injustice while simultaneously making injustice less detectable. Developers create algorithms with faulty inputs and flawed conclusions. They fail to test their models and rely on proxies that foster and often exacerbate discrimination. They create powerful engines that monetize attention, surveil consumers, and manipulate behavior without regard for the societal consequences. Their
deployment of algorithms also imperils competition. If left unaddressed, these algorithmic flaws will repeatedly and systematically harm consumers. These harms are often felt most acutely by already vulnerable or historically disadvantaged populations, especially Black
Americans and other communities of color.