Evaluating Racial Bias in AI Language Models [Poster]

Öffentlichkeit Deposited

Creator
Mitwirkende
Abstract
  • As AI models like ChatGPT become more prevalent in industries, my study investigates whether they replicate the racial biases often found in human decisions. This research examines ten AI models that are given the role of job recruiters, evaluating six resumes with nearly identical qualifications that differ only in the applicants' names, suggestive of Hispanic, Black, or White ethnic groups. Each AI model scores the resumes on a 0-10 scale. The initial findings reveal varying levels of bias; some AIs show substantial bias, while others demonstrate little to none. Incorporating the latter could help reduce racial prejudice in hiring. My research not only uncovers potential biases in AI but also offers insights into promoting fair AI use in the job market.

Language
Stichwort
Date created
Resource type
Rights statement

MLA citation style (9th ed.)

Jessup, Matthew. Evaluating Racial Bias In Ai Language Models [poster]. Burnette, Joyce.. 2024. wabash.hykucommons.org/concern/generic_works/4730e819-6435-4a78-8ca2-21a6931b5885?locale=de.

APA citation style (7th ed.)

J. Matthew. (2024). Evaluating Racial Bias in AI Language Models [Poster]. https://wabash.hykucommons.org/concern/generic_works/4730e819-6435-4a78-8ca2-21a6931b5885?locale=de

Chicago citation style (CMOS 17, author-date)

Jessup, Matthew. Evaluating Racial Bias In Ai Language Models [poster]. 2024. https://wabash.hykucommons.org/concern/generic_works/4730e819-6435-4a78-8ca2-21a6931b5885?locale=de.

Note: These citations are programmatically generated and may be incomplete.