AI-Powered Resume Searching in Government Contracting: Increasing Accuracy and Convenience with OpenAI; Artificial Intelligence in Job Recruitment: Navigating Integrating AI Ethically

At my internship at a government contracting company, we saw inefficiency and inaccuracy in the process that assigns employees to new government project proposals. To solve this, I built an application that filters and searches through employee resumes with Artificial Intelligence (AI) generated keywords to recommend the best. This application is detailed in my Technical Project. Throughout the development of this application, the main conversations with my team consisted of discussing methods to maintain data privacy and fairness between all employees. These discussions lead me to research the ethical considerations of artificial intelligence in job recruitment for my STS Research Project.
In my Technical Project paper, I detailed the design, development process, and results of the AI-powered resume analysis application I built during my internship. This application allows a user to filter employees by specific requirements, like years’ experience or security clearance level, and search for the most qualified employee by searching for the specific topic of the new government project proposal, like ‘hypersonics’, or specific similar past government project. The application has five key components: user input and interface, accessing resumes and data in real-time, filtering employees by specific requirements, searching employee resumes using AI-generated words, and displaying results. We used OpenAI’s API to generate a larger set of words and projects that represent the user input search word and past project. Using these words, the application searches through employee resumes and suggests the employees with the most relevant experience to the specific search. This project resulted in a ‘proof of concept’ application that received a positive response from administration and resulted in an increase in efficiency and accuracy of employee recommendation, but still maintains data privacy and fairness.
In my STS Research paper, I research and discuss instances of artificial intelligence used in job recruitment that show ethical concerns and negative impacts on applicants. Specifically discussing how proxy variables, biased training data, and automated video interviews can reproduce and reinforce discrimination and biases. I apply UNESCO’s Recommendation on the Ethics of Artificial Intelligence as an ethical framework to discuss, approach, and propose a solution to the ethical concerns of AI in job recruitment. Specifically looking at fairness and non-discrimination, right to privacy and data protection, human oversight and determination, responsibility and accountability, and transparency and explainability. My paper acknowledges that to gain the benefits of efficiency, cost reduction, and decrease in human bias that should come with the integration of AI in job recruitment, we need to establish an ethical framework, put policy in place, and establish proper testing to eliminate the negative impacts on vulnerable applicants such as unfairness, discrimination, lack of transparency, and more.
Job recruitment can drastically affect an applicant’s life, so when considering integrating artificial intelligence into the process it is important to acknowledge the ethical implications throughout the development process and ensure the design does not produce negative impacts. During my Technical Project, we had many discussions throughout the application’s development to ensure it maintained our data privacy and fairness ethical standards, specifically with its use of AI. My STS Research dives deeper into the ethical implications of the technology I implemented in my Technical Project on a larger scale, a technology that impacts applicants globally due to the increased use of AI in job recruitment. I learned that although this technology is widely used, it is also widely unchecked. There needs to be more conversations about the ethics of AI and more policies put in place to ensure applicants are treated fairly and ethically.
I would like to acknowledge and thank my STS Advisor, Professor Richard Jacques, and my Technical Advisor, Professor Brianna Morrison. I would also like to acknowledge my Capstone professor, Professor Rosanne Vrugtman. Additionally, I would like to acknowledge my Technical Project partner Kevin Hallissey, a student at Virginia Tech.

School of Engineering and Applied Science

Bachelor of Science in Computer Science

Technical Advisor: Briana Morrison

STS Advisor: Richard Jacques

Technical Team Members: NA