Mitigating Bias in Retail AI Hiring Tools: A Human-Centered H-A-I-R-E Governance Framework for Low-Wage Hires

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

AI reduces time, human effort, and manual errors.

“Although AI reduces time, human effort, and manual errors [4], but ethical problems occur, especially when it is not well designed [5].”

• cites [4] [4]. Bhargava, A., Bester, M., & Bolton, L. (2021). Employees' perceptions of the implementation of robotics, artificial intelligence, and automation (RAIA) on job satisfaction, job security, and employability. Journal of Technology in Behavioral Science , 6 (1), 106-113. - sourceSupported

May combine two claims.

Without transparency in AI decision-making processes has raised concerns about accountability, responsibility, ethics, and fairness as management may not be fully aware of decision-making.

“Moreover, without transparency in AI decision-making processes has raised concerns about accountability, responsibility, ethics, and fairness as the management may not be fully aware of the decisionmaking [10].”

• cites [10] [10]. Addis, C., & Kutar, M. (2019, August). AI management an exploratory survey of the influence of GDPR and FAT principles. In 2019 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI) (pp. 342-347). IEEE. - sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing).

Post acceptance of AI in HRM, the application processing system including screening, shortlisting, interview scheduling, and offer roll-out in retail has become efficient and effective.

“Post acceptance of artificial intelligence (AI) in HRM, the application processing system, which includes screening, shortlisting, interview scheduling, and offer roll-out in the retail sector, has become efficient and effective [1].”

• cites [1] [1]. Afolabi, J. O. A., Olatoye, F. O., Eboigbe, E. O., Abdul, A. A., & Daraojimba, H. O. (2023). Revolutionizing retail: hr tactics for improved employee and customer engagement. International Journal of Applied Research in Social Sciences, 5(10), 487514. - sourceCould not verify

  The section on Recruitment Patterns and Policy Adaptation in Retail HR discusses adapting recruitment policies, while the discussion of technological influences mentions streamlining HR transactions. Neither addresses AI-driven screening, shortlisting, interview scheduling, or offer roll-out, or establishes that this application process became more efficient and effective. Searches for those steps and for AI in HR recruitment found no supporting passage in our copy; page numbers were unavailable. (LLM, conf=0.73)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim. May combine two claims.

AI systems become smarter, independent, and take decisions which may create prejudiced or harmful behavior, and managers don't know how to apply ethics in decision-making.

“AI systems become smarter, independent, and take decision which may create prejudiced or harmful behavior, and managers don't know how to apply ethics in decision-making [13].”

• cites [13] [13]. Brendel, A. B., Mirbabaie, M., Lembcke, T. B., & Hofeditz, L. (2021). Ethical management of artificial intelligence. Sustainability , 13 (4), 1974. - sourcePartly supported

  The Introduction and Section 2 describe evolving AI with increasing decision-making autonomy and potentially harmful consequences; Section 3 gives an example of AI disadvantaging job applicants. The Introduction says ethical considerations need to inform managerial decisions and that a fragmented discourse is difficult for some employees to navigate. It does not establish the broader claim that managers do not know how to apply ethics in decision-making. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Ethical AI needs different approaches that integrate human involvement, accountability, and management intervention.

“The scholarly argument is that ethical AI needs different approaches that integrate human involvement, accountability, and management intervention [14].”

• cites [14] [14]. Zhou, J., Chen, F., Berry, A., Reed, M., Zhang, S., & Savage, S. (2020, December). A survey on ethical principles of AI and implementations. In 2020 IEEE Symposium Series on Computational Intelligence (SSCI) (pp. 3010-3017). IEEE. sourceSupported

May combine two claims.

It is difficult for hiring managers to explain selection/rejection when there is no clear accountability, which may reduce trust in AI tool and point out need for ethical governance systems.

“It is difficult for hiring managers to explain the reason for the selection or rejection of the candidate when there is no clear accountability, which may reduce trust in the AI tool and point out the need for ethical governance systems [11].”

• cites [11] [11]. Baird, A., & Schuller, B. (2020). Considerations for a more ethical approach to data in AI: On data representation and infrastructure. Frontiers in big Data , 3 , 25. - sourcePartly supported

  Section 4.4 (page unavailable) says that explainability concerns why AI decisions were made and that, without an explanation, results are harder to substantiate. Sections 4.3 and 7 also discuss trust and ethical considerations. The paper does not establish the specific claim about hiring managers explaining candidate selection or rejection when accountability is unclear. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Ethical problems occur in AI hiring especially when it is not well designed.

“Although AI reduces time, human effort, and manual errors [4], but ethical problems occur, especially when it is not well designed [5].”

• cites [5] [5]. Kelan, E. K. (2024). Algorithmic inclusion: Shaping the predictive algorithms of artificial intelligence in hiring. Human Resource Management Journal , 34 (3), 694-707. - sourceSupported

May combine two claims.

Mitigating technical bias is not sufficient and requires a structured governance framework.

“This research highlights the fact that mitigating technical bias is not sufficient and requires a structured governance framework [7].”

• cites [7] [7]. Yanamala, K. K. R. (2023). Transparency, privacy, and accountability in AI-enhanced HR processes. Journal of Advanced Computing Systems , 3 (3), 10-18. - sourceSupported

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

AI helps in matching resumes with job description based on skills and speeds up hiring process.

“AI helps in matching resumes with the job description based on the skills and speeds up the hiring process [8].”

• cites [8] [8]. Gupta, A., & Mishra, M. (2022). Ethical concerns while using artificial intelligence in recruitment of employees. - sourcePartly supported

  The Literature Review says AI can screen CVs and save recruiters time, supporting the screening and speed aspects of the claim. It does not say that AI matches resumes to job descriptions based on skills; the passages found describe screening and selecting candidates more generally. (LLM, conf=0.87)

  Experimental — accuracy not yet measured

May combine two claims.

The H-A-I-R-E engineering framework was developed based on the study's findings to support human review, audits, inclusive data practices, and ethical hiring decisions.

“Based on these findings, the H-A-I-R-E engineering framework (Human Oversight, Accountability, Inclusivity, Responsibility, and Ethics) was developed to support organizations in maintaining human review, conducting audits, ensuring inclusive data practices, and promoting ethical hiring decisions.”

May combine two claims.

Only 30.4% mention Yes to recruitment tool fairly evaluate candidate with gaps, 43.5% Maybe, 26.1% No.

“The research data shows only 30.4% mention 'Yes' to the recruitment tool fairly evaluate candidate with gaps in employment or education, but 43.5% say 'Maybe', and 26.1% 'No'.”

May combine two claims.

Majority of organizations are using AI tools to evaluate applications at resume screening (82.6%) and shortlisting (47.8%) stage, and lesser usage of conducting video interviews (26.1%) and offer roll-out (17.4%).

“The majority of organizations are using AI tools to evaluate applications at the resume screening (82.6%) and shortlisting (47.8%) stage, and lesser usage of conducting video interviews (26.1%) and offer roll-out (17.4%).”

May combine two claims.

21.7% mention Yes, 52.2% Not Sure due to lack of awareness or clarity, and 26.1% say No.

“21.7% mention 'Yes', and 52.2% indicate 'Not Sure' due to lack of awareness or clarity, and 26.1% say 'No'.”

May combine two claims.

39.1% sees AI as unfair, 13% very fair, 47.8% fair.

“A large participant 39.1% sees AI as unfair, with 13% express very fair, and 47.8% say fair.”

May combine two claims.

Less than 5% say recruiters and HR managers have AI decision override power, 43.5% can override Rarely, 17.4% Not Sure, 34.8% Sometimes.

“Moreover, less than 5% say recruiters and HR managers have AI decision override power, 43.5% can override AI decisions 'Rarely', 17.4% are 'Not Sure', and 34.8% state 'Sometimes'.”

May combine two claims.

Low transparency appears where 52.2% indicate organization clearly communicates technology used to evaluate resumes, and 47.8% say No/Not Sure, almost half.

“Low transparency appears that where 52.2% indicate that the organization clearly communicates that the technology is used to evaluate resumes, and 47.8% say 'No' and 'Not Sure', which is almost half of the participants.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Total sample size n=23 responses were received from multiple stakeholders.

“A total sample size of n=23 responses were received from multiple stakeholders i.e. HR & Recruitment managers, Store or Frontline managers, frontline or customer-facing employees, and entry-level workers.”

Classified as probably not a substantial claim. May combine two claims.

Job seekers always want human oversight when profiles are in hiring process, and they don't trust outcome if oversight missing.

“Job seekers always want human oversight when their profiles are in the hiring process, and they don't trust the outcome if oversight is missing in the hiring process.”

May combine two claims.

Respondents suspect real bias, and candidates are vocal when they aren't satisfied with result.

“The respondents suspect real bias, and candidates are vocal when they aren't satisfied with the result.”

May combine two claims.

Only 17.4% believe HR managers double-check resumes rejected by recruitment tools, 34.8% Yes in some cases, 34.8% No, 13% Not Sure.

“The study says that only 17.4% believe that HR managers double-check the resumes that are rejected by the recruitment tools, but 34.8% state 'Yes, in some cases', 34.8% mention 'No', and 13% 'Not Sure'.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Usage for AI recruitment tool for low-wage or entry-level hires is high, where more than 72% employers are using technology-based recruitment tools to validate resumes.

“Usage for AI recruitment tool for low-wage or entry-level hires is high, where more than 72% employers are using technology-based recruitment tools to validate resumes.”

Classified as probably not a substantial claim.

The H-A-I-R-E engineering framework was developed in response to the gap in research on governance practices and stakeholder participation.

“The H-A-I-R-E engineering framework was developed in response to this gap.”

May combine two claims.

33.3% are Fully Aware of AI hiring system makes decisions, and 66.67% are Not Aware/Not Sure.

“The study also observed that 33.3% are 'Fully Aware' of AI hiring system makes decisions, and 66.67% are 'Not Aware' and 'Not Sure' which showcases low process clarity.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

56.5% HR should be responsible when AI decision is incorrect, 17.4% Vendor, 4.4% Hiring Manager, 21.7% Not Sure.

“The study indicates that 56.5% HR should be responsible when AI decision is incorrect, 17.4% mention 'Vendor', 4.4% state 'Hiring Manager', and 21.7% say 'Not Sure'.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

For existing governance framework, 43.5% believe structured AI governance framework is in place, 43.5% Maybe, 13% No.

“For the existing governance framework, 43.5% believe a structured AI governance framework is in place, but 43.5% say 'Maybe', 13% indicate 'No'.”

Classified as probably not a substantial claim.

21.7% Strongly Agree, 26.1% Agree, 47.8% Somewhat Agree that ethical governance framework reduces bias in low wage retail hiring.

“A substantial proportion of the stakeholders highlight (21.7% 'Strongly Agree', 26.1% 'Agree', and 47.8% 'Somewhat Agree') ethical governance framework reduces bias in low wage retail hiring.”

May combine two claims.

Majority of participants think human involvement is essential for final hiring decisions. Overall 65.2% Yes always, 21.7% Yes in some cases, 13% Not Sure.

“The study also highlights that the majority of participants think that human involvement is essential for final hiring decisions. Overall, 65.2% state 'Yes, always', 21.7% say 'Yes, in some cases', and 13% mention 'Not Sure'.”

Classified as probably not a substantial claim. May combine two claims. The quote only partly states this claim.

Respondent highlights importance of Structured AI Framework: Human Oversight (47.8%), Accountability (56.5%), Inclusivity (30.4%), Responsibility (34.8%), Ethics (52.2%).

“In the study, the respondent's highlights an importance of a Structured AI Framework”

Classified as probably not a substantial claim.

Study highlights ethical and governance challenges in AI tools for low-wage and entry-level hires in retail companies.

“The study highlights ethical and governance challenges in AI tools for low-wage and entry-level hires in retail companies.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

As per participants opinion, 26.1% regular audits happen, 17.4% No audit, 56.5% Not Sure.

“As per the participants opinion, 26.1% mention regular audits happen, 17.4% 'No' audit takes place, and 56.5% are 'Not Sure'.”

May combine two claims.

AI hiring creates ethical problems including bias, transparency issues, and risk to candidate personal information.

“As per [6], AI hiring is creating ethical problems. It creates bias, transparency issues, and risk candidate's personal information.”

• cites [6] [6]. Mahapatra, T. (2025). AI and HR: Exploring the Ethical Implications of Algorithmic Hiring. Training and Development Journal , 16 (1), 1-13. - sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S2 (checkable).

AI hiring tools are sociotechnical systems.

“Algorithms-centric hiring tools are engineered sociotechnical systems.”

• Also stated in: § I. Introduction: · p. 3 · plain

  “The AI hiring tool are sociotechnical system because they combine technology and social processes like human oversight.”

Probably not a substantial claim: may not pass S3 (load-bearing).

AI has been introduced to the retail sector to improve recruitment and other HR areas.

“The AI has been introduced to the retail sector to improve the performance of recruitment and other HR areas [2].”

• cites [2] [2]. Rezvi, R. I., Rahman, K. O., Nasrullah, F., Islam, M. S., Hasan, M., Nusrat, N., ... & Ahmed, S. (2025). The integration of artificial intelligence in human resource management in the US retail sector. Journal of Business and Management Studies , 7 (1), 273-278. - sourceNot checked — no full text of the source

Probably not a substantial claim: may not pass S3 (load-bearing).

HR professionals use AI to screen and shortlist profiles for quick turnaround.

“The human resource professionals are using AI to screen and shortlist the profiles for quick turnaround [3].”

• cites [3] [3]. Black, J. Stewart, and Patrick van Esch. "AI-enabled recruiting: What is it and how should a manager use it?." Business horizons 63, no. 2 (2020): 215-226. - sourceNot checked — no full text of the source

May combine two claims.

A human-centered framework such as H-A-I-R-E may reduce discrimination, build trust among candidates, and promote fairness by controlling AI tools, especially for low-wage hires.

“The research shows that a human-centered framework, such as H-A-I-R-E, may reduce discrimination, build trust among candidates, and promote fairness by controlling AI tools, especially for low-wage hires.”

• AssumptionRests on an assumption: Human oversight, accountability, inclusivity, responsibility, and ethics are core design requirements for responsible AI hiring tools.

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

These safety measures ensure that it mitigates risk of discriminatory outcomes.

“These safety measures ensure that it mitigates the risk of discriminatory outcomes.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

This participation helps engineers learn to anticipate and resolve bias as a core design need.

“This participation helps engineers learn to anticipate and resolve bias as a core design need.”

Probably not a substantial claim: may not pass S4 (based).

H-A-I-R-E framework is an operational framework instead of aspirational.

“The H-A-I-R-E engineering framework is an operational framework instead of aspirational.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Integrating H-A-I-R-E framework with engineering system design may prepare engineers to incorporate its core value system directly into technology framework.

“Integrating the H-AI-R-E framework with engineering system design may prepare engineers to incorporate its core value system directly into the technology framework.”

May combine two claims.

As AI hiring tool becomes essential part of recruitment process in retail, there is growing risk perpetuate bias, compromising fairness, and low transparency.

“As the AI hiring tool becomes an essential part of the recruitment process in retail, there is a growing risk perpetuate bias, compromising fairness, and low transparency.”

Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

Overall, this research highlights that AI hiring tool needs a sociotechnical framework to be an ethical and fair structure design that make ready upcoming students and engineers to build trusted, responsible, and socially accountable AI systems.

“Overall, this research highlights that AI hiring tool needs a sociotechnical framework to be an ethical and fair structure design that make ready upcoming students and engineers to build trusted, responsible, and socially accountable AI systems.”

May combine two claims.

Findings show indicative insight that limited human oversight, opaque transparency, accountability, and trust in automation tools may cause unfair outcomes.

“Findings also show indicative insight that limited human oversight, opaque transparency, accountability, and trust in automation tools may cause unfair outcomes.”