Rose Bialk/The Mossy Log
Your resume was rejected in under six seconds. Not by a recruiter. Not by a hiring manager. But by an algorithm.
Before a human ever considers your experience or glances at your resume, a machine decides whether you are worth considering at all. This is not science fiction or a distant future — these are the mechanisms of hiring in the modern day. Applicant tracking systems now filter applications, rank candidates and quietly determine who gets access to an opportunity. The artificial intelligence that is embedded in these systems is often framed as a neutral and efficient technological fix for flawed human judgment, promising standardized objectivity that humans are unable to achieve.
History suggests that we be skeptical of that promise, since technologies do not exist in isolation from the systems that produce them. AI learns from data shaped by decades of exclusion, inequality and biased decision-making. Therefore, when we allow these historical patterns to be treated as benchmarks for “fitness” or “success,” discrimination is not eliminated, but automated within the system.
For a long time, the dominant fear surrounding AI has been job loss. The World Economic Forum’s Future Jobs Report for 2025 fuels this anxiety, noting that 40% of employers anticipate reducing their workforce in areas where AI can automate tasks. While this concern is real and deserves attention, I believe the current integration of AI tools in the hiring market demands immediate discourse. After watching the documentary “Coded Bias”in my Digital Media & Society class, I believe that the entrenched inequalities orchestrated through these tools are a more urgent issue than the replacement of our jobs. This documentary — which exposes how algorithmic systems already harm marginalized communities — makes it evident that job replacement is only a part of the bigger story. AI is already reshaping access to work, housing, healthcare and surveillance in ways that disproportionately harm women, people of color and those pushed to the margins. The impact is happening now, and it demands urgent scrutiny.
Hiring is one of the clearest domains where this harm is visible. Algorithmic decision-making is no longer experimental or marginal; it has become the infrastructure of several companies, such as Amazon, Unilever, Chipotle and Goldman Sachs. For students and early career workers, this often means their first interaction with the labor market is not a conversation but a filter. According to a 2026 report from the data analytics platform Demandsage, 87% of companies worldwide now utilize AI-powered tools in hiring, resulting in an AI recruitment industry valuation of $661.56 million in 2023. Experts also project this number to surpass $1.12 billion by 2030.
The same report highlights that over 65% of recruiters report using AI to save time, improve candidate sourcing and reduce hiring costs by as much as 30% per hire. Here, this technology is marketed with efficacy as the selling point. The question is, though, “Efficacy for whom?”
When algorithms are trained on biased data, they normalize these biases. A well-known example of this is Amazon’s now-scrapped AI recruiting tool. According to Reuters, by 2015, the company’s new system stopped rating candidates for software developer jobs and other technical posts in a gender-neutral way. The algorithm was trained on a decade’s worth of resumes, mostly from men. Engineers later admitted that the algorithm “did not like women,” not because it was explicitly programmed to discriminate, but because it learned from patterns in a male- dominated industry.
While one might think these discriminatory concerns with giant companies such as Amazon would have led to strict legislation or stirred up impactful conversations around the ethicality of tools, the future trends say otherwise. Amazon abandoned the tool after backlash in 2015, but the logic behind it remains embedded incountlesshiringsystemsstillinuse today. In 2023, a class action lawsuit was filed against Workday, alleging that the company had violated federal and state anti-discrimination laws through its AI-driven hiring tools. The plaintiff, Derek Mobley, alleged that Workday’s applicant screening systems, including automated assessments and personality tests, systematically discriminate, or enabled employers to discriminate against applicants based on race, age and disability. A trial court recently rejected Workday’s attempt to dismiss the case, allowing several claims to move forward. The lawsuit serves as a critical warning: legal accountability does not stop with software developers, but extends to the employers who deploy these systems.
Cases like these have not stopped, and are not expected to stop either. A 2024 study by the University of Washington Information School found that widely used AI resume screening tools disproportionately favored White and male candidates, with resumes linked with White-sounding names preferred 85% of the time. This consistency across models from different developers suggests these issues are widespread rather than company-specific.
What makes AI especially dangerous as a gatekeeping technology is its opacity. When a human hiring manager discriminates, biases can be identified, challenged and addressed. But when an algorithm does the same, exclusion is reframed as a technical outcome rather than political bias. Applicants are rejected without explanation, decisions are justified through proprietary models and automated rankings cannot be audited. Bias is invisible, automated and nearly impossible to contest.
This disproportionately harms already disadvantaged communities. Applicants from underfunded schools, non-traditional backgrounds or marginalized identities are more likely to submit resumes that deviate from “successful” profiles. AI systems flag those deviations as risk. Instead of opening doors, automation narrows access, filtering out people who do not resemble those already inside.
AI is not destiny, it is infrastructure. And infrastructures reflect choices about who is prioritized, who is excluded and whose labor is valued. In “Coded Bias,” Joy Buolamwini reminds us that power lies with those who own the code. If left unexamined, AI will continue a long traditionoftechnologicaladvancement that benefits the powerful while marginalizing those already on the edge of opportunity.
Hence, for students entering this labor market, panic is understandable but misplaced. Competing with machines on speed, keyword optimization or pattern recognition is a losing game. What is needed instead is critical AI literacy. Students must understand how algorithms are trained, where bias enters datasets and how automated systems shape access to opportunity. This is not merely about technical skills, but civic ones as well. The real danger is not that machines will replace us, but that we will allow them to decide which humans are worth considering long before a human ever gets the chance to disagree.
About the Author
