© July 2025| IJIRT | Volume 12 Issue 2 | ISSN: 2349-6002

Artificial Intelligence in Recruitment: A Critical Study of

Amazon’s AI Hiring Tool

Bidhan Kumar Rout

Student, [Link]. (Hons), Hansraj College, Delhi Affiliated to University of Delhi Delhi India

Abstract - Many companies are making a strong organizations, companies like Amazon are
investment into the future of AI to optimise their continuously looking to improve the hiring process. In
operations in the age of cognitive systems. A remarkable
order to manage new hires and keep costs down, most
case in point was witnessed in 2014, as Amazon built an
AI recruiting tool by utilizing Natural Language companies are starting to implement AI (Artificial
Processing (NLP) and Machine Learning (ML). The goal Intelligence) powered solutions.
of this system was to automate the process of screening
job applications and to identify best candidates by In 2014, Amazon had reportedly formed a secret team
analyzing resumes. It leveraged cutting-edge text to build an AI-driven hiring tool with a primary focus
analysis to parse out critical information and provide to streamline the resume screening procedure. This
insights to guide hiring. The system had initially been
project was built with NLP Machine Learning to
hailed as an encouraging step toward eliminating human
bias and making hiring more efficient. But in practice the classify candidates resumes and calculate ratings
system was found to be inadvertently discriminating which explained under what position they would fit in.
against them. This was an AI model that started favoring This system included a five star scale and was rapidly
male employees (particularly if they were applying for adopted across the organisation for its efficacy in
technical positions), simply because it had been trained
processing large numbers of applications.
on historical hiring data, and hiring history had been
sexist. This has raised significant ethical questions of
fairness and accountability in automated decision But a closer examination of the tool’s results led to
making. Due to these problems, Amazon pulled the plug some concerns that the tool could have some bias built
on the project. The episode underscored the dangers of into its algorithm. It was noted that the AI trained on
using AI in sensitive areas such as hiring, and it offered a historical hiring data was already starting to inherit
warning to business leaders about the necessity of
bias—specifically that more men were being
unbiased data and ethical AI development.
Consequently, Amazon opted to discontinue the AI-based recommended for technical jobs. It ended up shining a
hiring system. spotlight on a major shortcoming of the system: the
Keywords: artificial intelligence, gender bias, technology,
need to address larger conversations over whether it’s
talent and recruiting, human resources, machine learning, ethical to entrust AI with managing human resources.
hiring, technical, automated
Due to these issues, Amazon eventually made the
INTRODUCTION decision to stop using the tool. This case illustrates the
necessity to ensure fairness, transparency and ethical
Amazon is known for its disruption of well-established oversight when implementing AI-technologies in
industries through technological innovation and mass sensitive areas such as recruiting.
scale. It is the world's largest online marketplace, AI
assistant provider, and cloud computing platform as LITERATURE REVIEW
measured by revenue and market capitalization.
Amazon was founded by Jeff Bezos in Bellevue, The application of AI in human resource management
Washington, on July 5, 1994. From its modest has attracted much attention in academia and industry.
beginning, the company has become one of the Traditional recruitment methods, as Y. Acikgoz (2019)
leading, most progressive companies. Hiring grows points out also (and are still) used because of how
with the company: As companies realize that the reliable they are in obtaining participants, they are still
human resources are key to become successful inefficient, expensive and lengthy. These processes

IJIRT 182710 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 3176

 © July 2025| IJIRT | Volume 12 Issue 2 | ISSN: 2349-6002

usually include manual pre-screening, interviews and Design: The study was exploratory; we were interested
testing, which may reduce the effectiveness, to map the consequences, limitations, and ethical
particularly when large-scale hiring is involved. dilemmas concerning the use of AI in HR.

As organizations started considering a digital Sources of Data: Scholarly papers, news articles,
transformation of their HR operations, AI became industry reports, and case studies on AI and
attractive for automating CV screening and enhancing recruitment.
candidate matching. Kodiyan (2019) pointed out the
ethical issues and technical difficulties in AI-enabled Data Collection Criteria: Meta-synthesis through
hiring, especially considering Amazon’s automated document and content analysis enabled the discovery
resume screening tools. “Algorithmic decisions may of patterns, trends and issues associated with AI-
reproduce prejudice that already exists in the data and enabled hiring.
that can be used in a way that generates discriminatory
practice if people let it," he said. Focus Area: Amazon’s 2014 AI recruitment tool, its
implementation, and discontinuation due to bias
Dattner et al. Legal and ethical HR of AI in hiring(AI) concerns.
in hiring by(2019) from the Harvard Business Review
focused on the legal and ethical aspects of AI in the Limitation: Only secondary data are used in the study,
hiring. They observed that AI systems might be guilty no primary data (e.g., interviews or questionnaires)
of accidentally breaking employment equality laws if gathered.
they were not optimally designed or otherwise
ANALYSIS/DISCUSSION
regulated. In the same context, Dastin (2018) noted
how Amazon had to kill the AI recruiting engine, TRADITIONAL HIRING
which essentially worked against women, to
underscore its importance for fairness and According to research by Y. Acikgoz in 2019, there
transparency. isn't a one-size-fits-all model that organizations stick
to when it comes to recruitment. Still, most companies
In addition, Mujtaba and Mahapatra (2019) discussed generally follow a similar pattern that kicks off with
the ethical use of AI in HR by organizations. They spotting a job opening that needs to be filled. Once a
emphasized that while AI delivers efficiency, vacancy is identified, a thorough evaluation takes
companies need to establish algorithmic place to figure out the qualifications, skills, and
accountability, particularly when deployed in sensitive experience necessary for the role. This assessment lays
applications like hiring. the groundwork for selecting the best recruitment
channel—whether that’s through internal promotions,
Together the studies illustrate the opportunity and the
employee referrals, job boards, or recruitment
danger associated with the application of AI in a
agencies. After choosing the recruitment source, the
recruitment context. Promising more efficient and
process moves on to the screening phase, where
scalable hiring, experts warn that without tight
resumes or CVs from potential candidates are
controls and measures to mitigate bias, AI could
reviewed to find those who meet the basic eligibility
perpetuate existing hiring disparities.
requirements. This is then followed by various
METHODOLOGY selection methods like written tests, group discussions,
and interviews, all designed to evaluate candidates
Qualitative secondary data analysis is utilized in this more comprehensively. As highlighted by B. Mueller
study in order to examine the application of AI in and J.R. Baum in their 2011 study, these steps are
recruitment specifically looking at the case of crucial for employers to pinpoint and select the right
Amazon’s AI-based recruiting instrument. candidate who fits the organization's needs and
culture. While this traditional approach has been
around for quite some time and has produced reliable
results, it’s faced growing criticism for being both

IJIRT 182710 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 3177

 © July 2025| IJIRT | Volume 12 Issue 2 | ISSN: 2349-6002

slow and costly. With multiple stages and the need for management is Natural Language Processing (NLP).
human involvement at every turn, the process can lead This technology helps analyze and pull meaningful
to delays and higher recruitment expenses. This has insights from text data. When it comes to recruitment,
led many organizations to look into more efficient, NLP allows companies to automatically scan and
tech-savvy alternatives, like using Artificial interpret resumes, pinpointing essential details like
Intelligence (AI), to streamline and enhance their skills, qualifications, and work history. For example,
recruitment practices. Amazon has harnessed NLP in its AI-powered
recruitment tool to sift through resumes and pull out
ARTIFICIAL INTELLIGENCE AS A HIRING relevant information, enabling the system to sort and
TOOL categorize candidates based on how well they match
specific job roles. To ensure these AI systems work
Artificial Intelligence (AI) is all about machines
accurately, they need a lot of data for training. By
mimicking human intelligence and behavior, covering
being exposed to thousands of resumes and job
things like learning, reasoning, and problem-solving.
descriptions, the algorithms learn to spot patterns and
Lately, we've seen a big uptick in how businesses are
figure out which candidates are the best fit for a role.
using AI, particularly in Human Resource
As A. A. Kodiyan (2019) points out, data is crucial for
Management. More and more companies are jumping
these systems to operate effectively and provide sound
on the bandwagon with recruitment automation tools,
hiring recommendations.
and studies show that about 55% of them have already
woven AI into their hiring processes. The main goal of
WHY ARE COMPANIES LIKE AMAZON
this shift? To make recruitment quicker, more
ADOPTING ARTIFICIAL INTELLIGENCE?
efficient, and budget-friendly. By automating tasks
like screening resumes, shortlisting candidates, and We’re living in a time where intelligent machines and
scheduling interviews, companies aim to lighten the
digital innovation are taking center stage. Artificial
load for HR professionals and speed up the time it
Intelligence (AI) is truly transforming how industries
takes to fill open positions. Plus, AI tools are built to
operate and how we go about our daily lives. It’s
tackle common human errors, like unconscious bias or amazing to see how quickly AI has progressed; it can
missing details in resume evaluations, which can skew now tackle complex tasks that used to take humans
the fairness and accuracy of traditional hiring
years of training and expertise to master. From
methods. These technologies rely on data-driven
handling financial transactions and analyzing legal
algorithms to promote more consistent and objective
documents to detecting fraud and automating
decision-making. Because of this, many organizations customer service, AI is stepping in to do it all. These
are hopeful that AI will not only streamline advancements not only boost productivity but also
recruitment but also boost the quality of hires by
lessen our reliance on human effort for those
pinpointing candidates whose skills and experiences
repetitive, data-heavy tasks. The AI market is booming
closely match what the job requires.
like never before, fueled by ongoing
enhancements in computing power, data access, and
USE OF AI BASED HIRING TOOL BY AMAZON
machine learning techniques. This surge has attracted
significant investments from the private sector, with
The idea of Artificial Intelligence (AI) has been a hot
both startups and established companies eager to
topic in both academic circles and tech discussions for
develop cutting-edge AI technologies. Major players
quite some time, and it’s found its way into various
like Google, Facebook, and Amazon are leading the
research fields. Yet, it’s really only in the past ten years
charge, setting up dedicated AI research labs,
that we’ve seen technology advance to a point where
recruiting top talent, and pouring billions into
AI can be effectively used in organizations (G. Tecuci,
expanding the possibilities of AI. These companies
2010). Thanks to better computational power and
aren’t just using AI to improve their own operations—
access to vast amounts of data, AI has turned into a
think personalized recommendations, targeted ads,
valuable resource for tackling real-world business
and supply chain efficiency—they’re also providing
challenges. One standout use of AI in human resource
AI tools and platforms for businesses and developers

IJIRT 182710 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 3178

 © July 2025| IJIRT | Volume 12 Issue 2 | ISSN: 2349-6002

worldwide. Their commitment highlights how crucial female candidates at a significant disadvantage, going
AI is becoming as a strategic focus for the future. against the core principles of equality and
Consequently, AI is emerging as a vital force for nondiscrimination in hiring. Moreover, Amazon's use
innovation, economic growth, and competitive edge in of this AI technology brings up serious concerns about
the global marketplace. the predictability of its outcomes. The biased behavior
stems from historical data that reflects existing gender
PROBLEM FACED BY AMAZON imbalances in the tech industry. Implementing an AI
system without fully understanding the risks tied to
Amazon's AI recruitment system, which was meant to biased data could be viewed as negligence, especially
streamline the resume screening process, was found to given the potential damage to both candidates and the
have a troubling gender bias, showing a clear company's reputation. This situation underscores the
preference for male candidates, especially in technical need for careful monitoring and evaluation of
positions like software development. This issue arose technological solutions, particularly those that affect
from the data used to train the system, which primarily employment, to avoid unintentional harm and legal
reflected resumes and hiring trends from Amazon's repercussions.
past—an era that was predominantly male. CONCLUSION
Consequently, the AI developed a bias towards male
applicants, even undervaluing resumes that included Efficient hiring practices are crucial for any
terms like “women’s” or originated from women-only organization because making poor recruitment choices
educational institutions. It also showed a preference can lead to serious financial and operational
for certain male-associated language, such as the term challenges. To make these processes smoother, many
“executed,” which appeared more frequently in the companies, including Amazon, have started using
resumes of male candidates. Once Amazon recognized Artificial Intelligence (AI) to cut down on the time and
this bias, they chose to scrap the AI system before it costs that come with traditional hiring methods. AI
could be rolled out across the company, although they systems are built to help sift through and assess a large
made it clear that recruiters didn’t depend solely on the number of resumes quickly, aiming to make hiring
AI’s recommendations. Human recruiters were still more efficient and less resource-heavy. However,
crucial in the final selection process, meaning the AI Amazon's experience serves as a cautionary tale about
was only there to help sort through resumes. the risks of depending on poorly trained AI systems.
Nevertheless, the AI's impact on the initial candidate The AI, which was trained on past hiring data, ended
pool raised alarms about the risks of bias in automated up reinforcing gender bias by favoring male
hiring methods. This situation highlights the critical candidates, particularly for technical positions. This
need for fairness in AI systems, particularly in high- situation highlights the need to ensure that AI systems
stakes areas like hiring, and emphasizes the are trained on diverse and representative data and are
importance for companies to regularly audit and refine rigorously tested. Without these precautions, AI can
their algorithms to avoid reinforcing existing societal unintentionally perpetuate existing biases, leading to
biases. unfair outcomes and possible legal issues. While AI
can definitely improve the recruitment process, it’s
LEGAL CONSIDERATION IN THIS CASE evident that human involvement is still vital. During
the Covid-19 pandemic, Amazon and many other
companies utilized AI to aid in hiring but didn’t rely
When it comes to Amazon's AI recruitment system,
on it entirely. Instead, AI was used as a tool for sorting
there are quite a few legal issues to consider, especially
resumes and conducting initial candidate screenings,
regarding discrimination and employment law. The
while human recruiters kept the final say. This
system has shown a bias towards male candidates,
balanced approach ensures that while AI can boost
which could potentially violate the Employment
efficiency, human judgment remains essential for
Equality Act that prohibits gender discrimination. By
making fair and informed hiring decisions. In
undervaluing resumes that include gendered language
summary, while AI has the potential to significantly
or come from women-only institutions, the AI has put
enhance hiring efficiency, its implementation needs to

IJIRT 182710 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 3179

 © July 2025| IJIRT | Volume 12 Issue 2 | ISSN: 2349-6002

be approached with caution. Companies must ensure

that AI systems are trained on fair and unbiased data
and that human oversight is part of the process. By
blending the strengths of AI with human insight,
organizations can develop a more effective, equitable,
and efficient hiring process.

REFERENCES

[1] Dastin, J. (2018). Amazon scraps secret AI

recruiting tool that showed bias against women.
In Ethics of Data and Analytics (pp. 296-299).
Auerbach Publications.
[2] Dattner, B., Chamorro-Premuzic, T., Buchband,
R., & Schettler, L. (2019). The legal and ethical
implications of using AI in hiring. Harvard
Business Review, 25.
[3] Mujtaba, D. F., & Mahapatra, N. R. (2019,
November). Ethical considerations in AI-based
recruitment. In 2019 IEEE International
Symposium on Technology and Society (ISTAS)
(pp. 1-7). IEEE.
[4] Kodiyan, A. A. (2019). An overview of ethical
issues in using AI systems in hiring with a case
study of Amazon’s AI based hiring tool.
Researchgate Preprint, 1-19.
[5] Lewis, N. (2018). Will AI remove hiring bias.
Strategic HR Review. Retrieved July, 20, 2020.
[6] Mujtaba, D. F., & Mahapatra, N. R. (2019,
November). Ethical considerations in AI-based
recruitment. In 2019 IEEE International
Symposium on Technology and Society (ISTAS)
(pp. 1-7). IEEE.

IJIRT 182710 INTERNATIONAL JOURNAL OF INNOVATIVE RESEARCH IN TECHNOLOGY 3180