Chapter 1: The Need for Oversight

In the U.S., there's currently a glaring absence of federal oversight to protect the public from potentially harmful algorithms. While we have regulatory bodies like the USDA ensuring the safety of our food, the FDA ensuring the safety of our medications, and the NHTSA ensuring safe road travel, the same cannot be said for the algorithms powering our online experiences.

Consider how we interact with platforms like Google or Amazon—what safeguards exist for the algorithms that shape our searches and recommendations? This void in oversight is particularly concerning.

To illustrate the risks posed by these algorithms, I reviewed 62 studies that reveal a variety of harms inflicted by tech companies, from predatory advertising on Google to misleading health product suggestions on Amazon. While regulatory frameworks are still in their infancy, researchers and journalists have undertaken commendable efforts to document these algorithmic risks.

As these algorithmic systems continue to influence our digital interactions, understanding their potential harms becomes increasingly critical. I will outline four primary categories of algorithmic harms identified in my literature review: discrimination, distortion, exploitation, and misjudgment.

The video titled "Search Engine Breakdown: Are Algorithms Racist and Sexist? | Full Film" explores the implications of algorithmic biases and discrimination in depth.

Section 1.1: Discrimination in Algorithms

Algorithmic systems often reflect and exacerbate societal issues. There’s a growing awareness among lawmakers and the public about the capacity for algorithms to perpetuate discrimination. Several key studies have laid the groundwork for this understanding.

In one landmark study from 2013, researcher Latanya Sweeney examined the advertisements generated by Google searches for names that sounded distinctly Black versus those that sounded White. Alarmingly, searches for Black-sounding names often yielded ads related to criminal records, while White-sounding names did not.

Similarly, Safiya Noble's pivotal research unveiled disturbing results when she searched for family-friendly content for her stepdaughter. The top search results were often sexualized, highlighting the troubling nature of algorithmic outputs.

Noble argues that online racial disparities must be addressed, as they are intertwined with the very fabric of how information is disseminated in the digital age.

Subsection 1.1.1: Continuing Issues

This ongoing discrimination in algorithms calls for immediate attention and reform.

Chapter 2: Exploitation and Misjudgment

Algorithms often exploit user data and content generated by individuals. The prevailing narrative that "big tech profits off your data" is substantiated by studies showing how platforms like Facebook obscure the ways they utilize sensitive user information for targeted advertising.

Moreover, many algorithmic systems rely heavily on user-generated content from platforms like Wikipedia, raising ethical questions regarding fair compensation for contributors.

Misjudgment also emerges as a significant concern; algorithms frequently misinterpret user interests, leading to irrelevant advertising. These inaccuracies could contribute to what some have termed a "digital advertising bubble," where advertisers may not receive the value they expect.

Conclusion: A Call to Action

The rise of new technologies inevitably introduces risks, as evidenced by the 62 studies examined. From discrimination to distortion, exploitation to misjudgment, these algorithms can have real-world implications.

Yet, as noted by Arundhati Roy, "another world is not only possible, she is on her way." By uniting as a collective of users, researchers, and advocates, we can advocate for systems that prioritize public welfare over corporate profit. The path ahead may be challenging, but together we can foster healthier interactions with our technology.