Rejected by Code: Algorithmic Bias in AI-Driven Recruitment and the Case for HR-IT Governance Frameworks

May combine two claims.

Testing large language models in a resume-screening task, Wilson and Caliskan found that resumes with White-associated names were significantly preferred in 85.1% of tests, while resumes with Black-associated names were preferred in only 8.6%.

“Testing large language models in a resume-screening task, Wilson and Caliskan [12] found that resumes with White-associated names were significantly preferred in 85.1% of tests, while resumes with Black-associated names were preferred in only 8.6%.”

• cites [12] sourceSupported

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

Raji et al. propose an end-to-end internal auditing framework that traces an AI system's behavior across its development lifecycle, with the aim of catching harmful outcomes before deployment.

“Raji et al. [5] propose an end-to-end internal auditing framework that traces an AI system's behavior across its development lifecycle, with the aim of catching harmful outcomes before deployment.”

• cites [5] sourceSupported

May combine two claims.

Algorithmic bias in hiring is real and largely traceable to data and design rather than malice.

“Taken together, this literature establishes two claims the remainder of the paper relies on: that algorithmic bias in hiring is real and largely traceable to data and design rather than malice, and that fixing it cannot be left to the technology alone.”

• cites [2] sourceNot checked — no full text of the source
• cites [3] sourceNot checked — no full text of the source
• cites [11] sourcePartly supported

  Section 2.1 describes discriminatory delivery of STEM job ads despite a gender-neutral intention, and Section 3 explains how biased data and algorithm-design choices produce biased outcomes. These passages support bias in employment-related algorithms without malice, but they do not establish that hiring bias is largely attributable to data and design. (LLM, conf=0.84)

  Experimental — accuracy not yet measured

Classified as probably not a substantial claim.

Raji et al. argue that closing the accountability gap requires deliberate internal auditing across a system's lifecycle, precisely the end-to-end scrutiny that unstructured deployment lacks.

“Raji et al. [5] argue that closing this accountability gap requires deliberate internal auditing across a system's lifecycle, precisely the end-to-end scrutiny that unstructured deployment lacks.”

• cites [5] sourceSupported

May combine two claims.

The NIST AI Risk Management Framework provides a general lifecycle process applicable to any AI system in any sector, but by design it does not address the specific division of labor in hiring, where HR procures a tool that IT deploys.

“Standardized risk frameworks such as the NIST AI Risk Management Framework provide a general lifecycle process applicable to any AI system in any sector, but by design they do not address the specific division of labor in hiring, where HR procures a tool that IT deploys [22].”

• cites [22] sourcePartly supported

  The AI RMF 1.0 introduction describes the framework as non-sector-specific and use-case agnostic, and its Core sets out functions that can be applied across the AI lifecycle. Appendix A distinguishes procurement from deployment tasks, but our copy does not discuss a hiring arrangement in which HR procures a tool and IT deploys it, or state that this division of labor was deliberately excluded. (LLM, conf=0.85)

  Experimental — accuracy not yet measured

May combine two claims.

Technical auditing frameworks specify how to inspect a system; Raji et al. are explicit that determining what to audit and why lies outside their scope.

“Technical auditing frameworks specify how to inspect a system; Raji et al. [5] are explicit that determining what to audit and why lies outside their scope ([5]: p. 38).”

• cites [5] sourceSupported

May combine two claims.

AI recruitment tools encode and amplify historical bias through proxy variables rather than explicit prejudice, a mechanism established in the literature and, in the disability context, now recognized by a federal court as a plausible basis for liability.

“AI recruitment tools encode and amplify historical bias through proxy variables rather than explicit prejudice, a mechanism established in the literature and, in the disability context, now recognized by a federal court as a plausible basis for liability.”

• cites [13] sourcePartly supported

  Split into 2 atoms — 0 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 1 unverifiable. Expand the claim to see per-atom verdicts.

  Experimental — accuracy not yet measured
• cites [6] Mobley v. Workday, Inc., 740 F. Supp. 3d 796 (N.D. Cal. 2024).https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.80.0.pdfNot checked yet

  Not checked in the free run.

• cites [7] Mobley v. Workday, Inc., No. 3:23-cv-00770-RFL, 2025 WL 1424347 (N.D. Cal. May 16, 2025). https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.128.0.pdfNot checked yet

  Not checked in the free run.

May combine two claims.

Bias is a property of the whole sociotechnical system, and deploying an AI hiring tool without accounting for that context is a governance gap, not merely a technical oversight.

“Bias is a property of the whole sociotechnical system, and deploying a tool without accounting for that context is a governance gap, not merely a technical oversight.”

• cites [13] sourcePartly supported

  The Introduction treats fairness as a property of social and legal systems, not of technical tools alone. Section 3 says an automated résumé screen must be understood alongside the behavior of hiring managers and candidates, supporting the need to account for context. The paper does not specifically characterize deploying such a tool without that context as a governance gap. (LLM, conf=0.78)

  Experimental — accuracy not yet measured

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

The Mobley claims spanned race, age, and disability, brought under Title VII, the ADA, the ADEA, and California law, and proceeded on a disparate impact theory: not that Workday intended to discriminate, but that its screening produces discriminatory outcomes.

“The claims spanned race, age, and disability, brought under Title VII, the ADA, the ADEA, and California law, and they proceeded on a disparate impact theory: not that Workday intended to discriminate, but that its screening produces discriminatory outcomes.”

May combine two claims.

The court permitted a disability claim to proceed under a proxy discrimination theory, accepting that a machine-learning system need not record disability status directly to discriminate on that basis; it may infer and screen out candidates through correlated signals such as lengthy employment gaps or patterns consistent with medical leave and recovery.

“Second, and most relevant here, the court permitted a disability claim to proceed under a proxy discrimination theory, accepting that a machine-learning system need not record disability status directly to discriminate on that basis; it may infer and screen out candidates through correlated signals such as lengthy employment gaps or patterns consistent with medical leave and recovery.”

• cites [6] Mobley v. Workday, Inc., 740 F. Supp. 3d 796 (N.D. Cal. 2024).https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.80.0.pdfNot checked yet

  Not checked in the free run.

• cites [7] Mobley v. Workday, Inc., No. 3:23-cv-00770-RFL, 2025 WL 1424347 (N.D. Cal. May 16, 2025). https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.128.0.pdfNot checked yet

  Not checked in the free run.

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

Because the litigation remains unresolved, the extension of liability to vendors is an emerging rather than settled doctrine; the agency question, the merits of the disparate-impact claims, and the scope of the preliminarily certified collective all remain open.

“Because the litigation remains unresolved, however, the extension of liability to vendors described here is an emerging rather than settled doctrine; the agency question, the merits of the disparate-impact claims, and the scope of the preliminarily certified collective all remain open.”

• cites [6] Mobley v. Workday, Inc., 740 F. Supp. 3d 796 (N.D. Cal. 2024).https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.80.0.pdfNot checked yet

  Not checked in the free run.

• cites [7] Mobley v. Workday, Inc., No. 3:23-cv-00770-RFL, 2025 WL 1424347 (N.D. Cal. May 16, 2025). https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.128.0.pdfNot checked yet

  Not checked in the free run.

May combine two claims.

The allegations in Mobley describe a system integrated as the first gate every applicant passed through, scoring and sorting candidates at scale without configuration review, outcome monitoring, or any mechanism to explain a rejection.

“A system was integrated as the first gate every applicant passed through; it scored and sorted candidates at scale; and it did so without configuration review, outcome monitoring, or any mechanism to explain a rejection.”

• cites [6] Mobley v. Workday, Inc., 740 F. Supp. 3d 796 (N.D. Cal. 2024).https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.80.0.pdfNot checked yet

  Not checked in the free run.

• cites [7] Mobley v. Workday, Inc., No. 3:23-cv-00770-RFL, 2025 WL 1424347 (N.D. Cal. May 16, 2025). https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.128.0.pdfNot checked yet

  Not checked in the free run.

May combine two claims.

Employers using Workday's platform relied on scores and recommendations, "strong," "good," "fair," "low," generated by a system whose internal workings they did not control and, in many cases, could not inspect.

“Employers using Workday's platform relied on scores and recommendations, “strong,” “good,” “fair,” “low,” generated by a system whose internal workings they did not control and, in many cases, could not inspect.”

• cites [6] Mobley v. Workday, Inc., 740 F. Supp. 3d 796 (N.D. Cal. 2024).https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.80.0.pdfNot checked yet

  Not checked in the free run.

• cites [7] Mobley v. Workday, Inc., No. 3:23-cv-00770-RFL, 2025 WL 1424347 (N.D. Cal. May 16, 2025). https://storage.courtlistener.com/recap/gov.uscourts.cand.408645/gov.uscourts.cand.408645.128.0.pdfNot checked yet

  Not checked in the free run.

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

The NIST AI Risk Management Framework (2023) offers organizations a structured, voluntary process for identifying and managing AI-related risks across a system's lifecycle.

“The third group is standardized risk governance, exemplified by the National Institute of Standards and Technology's AI Risk Management Framework (2023), which offers organizations a structured, voluntary process for identifying and managing AI-related risks across a system's lifecycle.”

• cites National Institute of Standards and Technology's AI Risk Management Framework (2023)Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

May combine two claims.

A vendor whose system actively scores, ranks, and effectively rejects candidates is performing a screening function traditionally carried out by the employer, and is therefore acting as the employer's agent rather than as a passive tool.

“Applied to hiring technology, the argument is that a vendor whose system does not merely store or transmit applications, but actively scores, ranks, and effectively rejects candidates, is performing a screening function traditionally carried out by the employer, and is therefore acting as the employer's agent rather than as a passive tool [21].”

• cites [21] Ajunwa, I. (2020) The Paradox of Automation as Anti-Bias Intervention. <i>Cardozo Law Review</i>, 41, 1671-1742.Not checked yet

  Not checked in the free run.

May combine two claims.

Existing governance frameworks tell organizations to audit, think in sociotechnical terms, and manage risk, but do not resolve who is accountable when a biased hiring tool is procured by one function and run by another, each reasonably assuming the other has addressed the problem.

“Existing frameworks tell organizations to audit, to think in sociotechnical terms, and to manage risk, but they do not resolve who is accountable when a biased hiring tool is procured by one function and run by another, each reasonably assuming the other has addressed the problem.”

• AssumptionRests on an assumption: The HR and IT functions are distinct organizational functions with separate responsibilities in the procurement and deployment of AI hiring tools.

May combine two claims.

Under the ADA, a qualification standard or selection criterion that screens out individuals with disabilities must be job-related and consistent with business necessity, and the employer must additionally show that the required performance cannot be achieved with reasonable accommodation.

“The ADA differs again: a qualification standard or selection criterion that screens out individuals with disabilities must be job-related and consistent with business necessity, but the employer must additionally show that the required performance cannot be achieved with reasonable accommodation [16].”

• cites [16] Americans with Disabilities Act of 1990, 42 U.S.C. § 12101 et seq (1990). https://www.law.cornell.edu/uscode/text/42/chapter-126Not checked yet

  Not checked in the free run.

May combine two claims.

Under Title VII, once a plaintiff establishes that a selection practice produces a disparate impact, the employer must demonstrate that the practice is job-related for the position in question and consistent with business necessity; the plaintiff may then prevail by identifying a less discriminatory alternative that the employer refused to adopt.

“Under Title VII, once a plaintiff establishes that a selection practice produces a disparate impact, the employer must demonstrate that the practice is job-related for the position in question and consistent with business necessity; the plaintiff may then prevail by identifying a less discriminatory alternative that the employer refused to adopt [14] [17].”

• cites [14] (1964) Civil Rights Act of 1964, Title VII, 42 U.S.C. §§ 2000e to 2000e-17. https://www.law.cornell.edu/uscode/text/42/2000eNot checked yet

  Not checked in the free run.

• cites [17] Griggs v. Duke Power Co., 401 U.S. 424 (1971). https://supreme.justia.com/cases/federal/us/401/424/Not checked yet

  Not checked in the free run.

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

The rulings in Mobley mark a legal inflection point extending potential liability from the employer to the technology vendor itself.

“Second, the rulings in Mobley mark a legal inflection point that extends potential liability from the employer to the technology vendor itself.”

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

Ajunwa argues that automation should not dissolve accountability, contending that creators of runaway algorithms should not be permitted to disavow their creations but must bear ultimate liability for harm wrought by their creations.

“Ajunwa [21] argues that automation should not dissolve accountability, contending that “the creators of runaway algorithms should not be permitted to disavow their creations” but must instead “bear the ultimate liability for any harm wrought by their creations” ( [21]: p. 1707).”

• cites [21] Ajunwa, I. (2020) The Paradox of Automation as Anti-Bias Intervention. <i>Cardozo Law Review</i>, 41, 1671-1742.Not checked yet

  Not checked in the free run.

May combine two claims.

A vendor whose model independently determines which applicants advance looks like a participant in the employment decision, opening the door to direct liability.

“A vendor whose model independently determines which applicants advance looks like a participant in the employment decision, and it is that characterization that opens the door to direct liability.”

Classified as probably not a substantial claim.

Because the Mobley claims span Title VII, ADEA, and ADA, and the age claim is the one on which preliminary collective certification was granted, a framework that applies a single uniform test across them would misstate the governing standard for the paper's most developed claim.

“Because the Mobley claims span all three statutes, and because the age claim is the one on which preliminary collective certification was granted, a framework that applies a single uniform test across them would misstate the governing standard for the paper's most developed claim.”

May combine two claims.

In one large employer survey, more than 90% of firms used automated systems to initially filter or rank candidates, including 94% for middle-skill roles and 92% for high-skill roles.

“in one large employer survey, more than 90% of firms used automated systems to initially filter or rank candidates, including 94% for middleskill roles and 92% for high-skill roles [10].”

• cites [10] Fuller, J.B., Raman, M., Sage-Gavin, E. and Hines, K. (2021) Hidden Workers: Untapped Talent. Harvard Business School Project on Managing the Future of Work and Accenture.Not checked yet

  Not checked in the free run.

May combine two claims.

Vendors' claims to have mitigated bias vary widely in rigor and transparency, with little external verification.

“Raghavan et al. [3] reinforce this empirically, finding that vendors' claims to have mitigated bias vary widely in rigor and transparency, with little external verification.”

• cites [3] sourceNot checked — no full text of the source

May combine two claims.

In Smith v. City of Jackson, the Supreme Court held that disparate-impact claims are cognizable under the ADEA but that the employer's defense is that the practice rests on reasonable factors other than age, a standard narrower than Title VII's business-necessity test.

“In Smith v. City of Jackson [18], the Supreme Court held that disparate-impact claims are cognizable un- der the ADEA but that the employer's defense is that the practice rests on reasonable factors other than age, a standard the Court expressly distinguished from, and described as narrower in reach than, Title VII's business-necessity test.”

• cites [18] Smith v. City of Jackson, 544 U.S. 228 (2005).https://supreme.justia.com/cases/federal/us/544/228/Not checked yet

  Not checked in the free run.

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

The unresolved boundary between HR and IT, surfaced by Mobley, is the gap the framework proposed in this paper is designed to close.

“That unresolved boundary, surfaced starkly by the Mobley case, is the gap the framework proposed in this paper is designed to close.”

• AssumptionRests on an assumption: The HR and IT functions are distinct organizational functions with separate responsibilities in the procurement and deployment of AI hiring tools.

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

Managing this exposure requires a joint HR-IT governance framework in which neither function can discharge its responsibility by deferring to the other, to the vendor, or to the algorithm.

“that managing this exposure requires a joint HR-IT governance framework in which neither function can discharge its responsibility by deferring to the other, to the vendor, or to the algorithm.”

• Also stated in: § 1.1. Background and Research Question · plain

  “Third, managing this risk requires a joint HR-IT governance framework for AI procurement and oversight, one in which neither func-tion can discharge its responsibility by deferring to the algorithm.”

• AssumptionRests on an assumption: The HR and IT functions are distinct organizational functions with separate responsibilities in the procurement and deployment of AI hiring tools.

May combine two claims.

The EEOC guidance is influential but does not carry the force of law, and it presumes the employer is the party responsible.

“That guidance is influential but does not carry the force of law, and it shares a built-in assumption with the statutes themselves: it presumes the employer is the party responsible.”

• cites [19] U.S. Equal Employment Opportunity Commission (2022) The Americans with Disabilities Act and the Use of Software, Algorithms, and Artificial Intelligence to Assess Job Applicants and Employees. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_americans-disabilities-act-and-use-software-algorithms-and-artificial-intelligence.pdfNot checked yet

  Not checked in the free run.

• cites [20] U.S. Equal Employment Opportunity Commission (2023) Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures under Title VII of the Civil Rights Act of 1964. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_select-issues-assessing-adverse-impact-software-algorithms-and-artificial.pdfNot checked yet

  Not checked in the free run.

May combine two claims.

The tool-versus-participant line is the pivot on which vendor liability turns, and it is the seam where the governance failure this paper identifies takes root: an employer that treats the system as a mere tool while the vendor treats itself the same way leaves the screening function accountable to no one.

“This tool-versus-participant line, which the next section shows the Mobley court drawing explicitly, is the pivot on which vendor liability turns, and it is precisely the seam where the governance failure this paper identifies takes root: an employer that treats the system as a mere tool while the vendor treats itself the same way leaves the screening function accountable to no one.”

• AssumptionRests on an assumption: The HR and IT functions are distinct organizational functions with separate responsibilities in the procurement and deployment of AI hiring tools.

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

The NIST framework is general by design, written to apply to any AI system in any sector, and does not address the specific division of labor in hiring where HR selects and procures a tool that IT deploys and maintains.

“Its limitation for present purposes is that it is general by design, written to apply to any AI system in any sector, and so it does not address the specific division of labor in hiring, where a human resources function selects and procures a tool that an information technology function deploys and maintains.”

• AssumptionRests on an assumption: The HR and IT functions are distinct organizational functions with separate responsibilities in the procurement and deployment of AI hiring tools.

May combine two claims.

The EEOC has stated that an employer may be liable under Title VII or the ADA for a vendor-built tool, and that employers may be held responsible for the actions of their agents, which may include software vendors, if the employer has given them authority to act on the employer's behalf.

“The EEOC has affirmed this directly, stating in its technical assistance that an employer may be liable under Title VII or the ADA for a vendor-built tool, and further that employers “may be held responsible for the actions of their agents, which may include entities such as software vendors, if the employer has given them authority to act on the employer's behalf” [19] [20].”

• cites [19] U.S. Equal Employment Opportunity Commission (2022) The Americans with Disabilities Act and the Use of Software, Algorithms, and Artificial Intelligence to Assess Job Applicants and Employees. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_americans-disabilities-act-and-use-software-algorithms-and-artificial-intelligence.pdfNot checked yet

  Not checked in the free run.

• cites [20] U.S. Equal Employment Opportunity Commission (2023) Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures under Title VII of the Civil Rights Act of 1964. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_select-issues-assessing-adverse-impact-software-algorithms-and-artificial.pdfNot checked yet

  Not checked in the free run.

May combine two claims.

A vendor that provides neutral infrastructure, executing criteria the employer alone specifies, looks like a tool and attracts little independent liability.

“A vendor that provides neutral infrastructure, executing criteria the employer alone specifies, looks like a tool and attracts little independent liability.”

May combine two claims.

The four-fifths rule is a rule of thumb rather than a bright line; the same guidance notes it can be inappropriate for very large samples and that courts sometimes rely on tests of statistical significance instead.

“The four-fifths rule is a rule of thumb rather than a bright line; the same guidance notes it can be inappropriate for very large samples and that courts sometimes rely on tests of statistical significance instead [20].”

• cites [20] U.S. Equal Employment Opportunity Commission (2023) Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures under Title VII of the Civil Rights Act of 1964. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_select-issues-assessing-adverse-impact-software-algorithms-and-artificial.pdfNot checked yet

  Not checked in the free run.

May combine two claims.

The threshold between acceptable score and rejection is not a technical default; it is a policy decision with direct disparate-impact consequences, because moving the cutoff changes which protected groups clear the bar.

“That threshold is not a technical default; it is a policy decision with direct disparate-impact consequences, because moving the cutoff changes which protected groups clear the bar.”

May combine two claims.

An HR function that relies on the vendor's marketing rather than independent evidence has substituted the vendor's judgment for its own and has not exercised oversight.

“An HR function that relies on the vendor's marketing rather than on independent evidence has not exercised oversight; it has substituted the vendor's judgment for its own.”

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

The adoption of AI screening is a legitimate response to real operational pressure, but it is not a delegation that relieves HR of accountability for who is hired and who is excluded.

“The adoption of AI screening is a legitimate response to real operational pressure, but it is not a dele-gation that relieves HR of accountability for who is hired and who is excluded.”

Classified as probably not a substantial claim.

Under the four-fifths rule described in EEOC guidance, a selection rate for one group that is less than 80% of the rate for the most-selected group may indicate adverse impact.

“Under the four-fifths rule described in EEOC guidance, a selection rate for one group that is less than 80% of the rate for the most-selected group may indicate adverse impact [20].”

• cites [20] U.S. Equal Employment Opportunity Commission (2023) Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures under Title VII of the Civil Rights Act of 1964. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_select-issues-assessing-adverse-impact-software-algorithms-and-artificial.pdfNot checked yet

  Not checked in the free run.

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

A system passing every conventional test of performance can still produce unlawful outcomes, and configuration and deployment choices are policy decisions, not merely technical ones.

“Equally, the technology function must recognize that a system passing every conventional test of performance can still produce unlawful outcomes, and that configuration and deployment choices are policy decisions, not merely technical ones.”

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

The literature was assembled by citation chaining rather than systematic review, so closely adjacent work may not be represented.

“Because the literature was assembled by citation chaining rather than systematic review, closely adjacent work may not be represented.”

May combine two claims.

From an operational standpoint the deployment in Mobley worked, but from a governance standpoint it was ungoverned; the failure was not that the system malfunctioned, but that its normal, functioning output was never treated as something requiring oversight.

“From an operational standpoint the deployment worked.From a governance standpoint it was ungoverned, and that distinction is the point: the failure was not that the system malfunctioned, but that its normal, functioning output was never treated as something requiring oversight.”

May combine two claims.

When an applicant is rejected, the organization often cannot explain why, because the decision lives inside a vendor model it cannot inspect.

“When an applicant is rejected, the organization often cannot explain why, because the decision lives inside a vendor model it cannot inspect.”

• AssumptionRests on an assumption: When an applicant is rejected, the organization often cannot explain why, because the decision lives inside a vendor model it cannot inspect.

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

The proposed framework's contribution is that it maps each governance task to an accountable function and defines joint checkpoints-procurement, configuration, monitoring, and review-at which HR and IT must act together.

“This framework's contribution is precisely that specificity: it maps each governance task to an accountable function and defines the joint checkpoints, procurement, configuration, monitoring, and review, at which HR and IT must act together.”

May combine two claims.

A tool that exercises screening judgment carries the employer's legal exposure, and the EEOC has confirmed that this exposure is not transferred to the vendor by contract.

“A tool that exercises screening judgment carries the employer's legal exposure, and the EEOC has confirmed that this exposure is not transferred to the vendor by contract [19] [20].”

• cites [19] U.S. Equal Employment Opportunity Commission (2022) The Americans with Disabilities Act and the Use of Software, Algorithms, and Artificial Intelligence to Assess Job Applicants and Employees. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_americans-disabilities-act-and-use-software-algorithms-and-artificial-intelligence.pdfNot checked yet

  Not checked in the free run.

• cites [20] U.S. Equal Employment Opportunity Commission (2023) Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures under Title VII of the Civil Rights Act of 1964. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_select-issues-assessing-adverse-impact-software-algorithms-and-artificial.pdfNot checked yet

  Not checked in the free run.

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

Responsibility for algorithmic hiring is shared and cannot be delegated, neither to the other function nor to the vendor nor to the algorithm itself, and organizations therefore need a governance framework that assigns joint HR-IT ownership across the full life of the tool.

“The central argument of this paper is that responsibility for algorithmic hiring is shared and cannot be delegated, neither to the other function nor to the vendor nor to the algorithm itself, and that organizations therefore need a governance framework that assigns joint HR-IT ownership across the full life of the tool.”

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

Managing the exposure from AI hiring requires a governance framework in which neither HR nor IT can discharge its responsibility by deferring to the other, to the vendor, or to the algorithm.

“And managing that exposure requires a governance framework in which neither the human resources function nor the information technology function can discharge its responsibility by deferring to the other, to the vendor, or to the algorithm.”

May combine two claims.

On the account the litigation presents, the technology performed as designed; what was missing was any person or function prepared to answer for its outcomes.

“On the account the litigation presents, the technology performed as designed; what was missing was any person or function prepared to answer for its outcomes.”

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

These procurement questions are HR's to ask.

“These are procurement questions, and they are HR's to ask.”

• AssumptionRests on an assumption: HR departments are the functional owners of the hiring process; they select the tools, define the criteria, and answer for the outcomes.

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

If liability cannot be outsourced, then oversight cannot be either.

“If the liability cannot be outsourced, then the oversight cannot be either.”

May combine two claims.

The EEOC's guidance lists concrete questions an employer should ask a vendor before purchase: whether the tool was tested for disparate impact, whether it makes disability-related inquiries, and whether its measured traits correlate with protected characteristics.

“The EEOC's guidance lists concrete questions an employer should ask a vendor before purchase: whether the tool was tested for disparate impact, whether it makes disability-related inquiries, whether its measured traits correlate with protected characteristics [19] [20].”

• cites [19] U.S. Equal Employment Opportunity Commission (2022) The Americans with Disabilities Act and the Use of Software, Algorithms, and Artificial Intelligence to Assess Job Applicants and Employees. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_americans-disabilities-act-and-use-software-algorithms-and-artificial-intelligence.pdfNot checked yet

  Not checked in the free run.

• cites [20] U.S. Equal Employment Opportunity Commission (2023) Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures under Title VII of the Civil Rights Act of 1964. https://data.aclum.org/storage/2025/01/EOCC_www_eeoc_gov_laws_guidance_select-issues-assessing-adverse-impact-software-algorithms-and-artificial.pdfNot checked yet

  Not checked in the free run.

May combine two claims.

In practice, configuration settings are frequently configured once during implementation, by technical staff optimizing for fit and throughput, and then left unexamined, with no defined owner accountable for their downstream effect on protected classes.

“In practice, such settings are frequently configured once during implementation, by technical staff optimizing for fit and throughput, and then left unexamined, with no defined owner accountable for their downstream effect on protected classes.”

• AssumptionRests on an assumption: In practice, configuration settings are frequently configured once during implementation, by technical staff optimizing for fit and throughput, and then left unexamined, with no defined owner accountable for their downstream effect on protected classes.

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

The analysis rests on a single case still in active litigation, so its rulings are preliminary and may yet change on further proceedings or appeal, and the four-fifths screen it relies upon is a rule of thumb rather than a definitive test of discrimination.

“The analysis also rests on a single case still in active litigation, so its rulings are preliminary and may yet change on further proceedings or appeal, and the four-fifths screen it relies upon is a rule of thumb rather than a definitive test of discrimination.”

Classified as probably not a substantial claim.

The most consequential and least examined gap in IT deployment is configuration governance.

“The most consequential and least examined of these gaps is configuration governance.”

• AssumptionRests on an assumption: Enterprise hiring platforms expose configurable settings that materially change who is screened out.

May combine two claims.

Vendor claims of a "bias-free" or "validated" tool are frequently accepted at face value, yet scholarship shows such claims vary widely in rigor and are rarely subject to external verification.

“Vendor claims of a “bias-free” or “validated” tool are frequently accepted at face value, yet the scholarship shows such claims vary widely in rigor and are rarely subject to external verification [3].”

• cites [3] sourceNot checked — no full text of the source

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

The contribution of the paper is to insist that the answer to who is accountable when an automated hiring system says no cannot be the algorithm, and to offer a framework for ensuring it is not.

“The contribution of this paper is to insist that the answer cannot be the algorithm, and to offer a framework for ensuring it is not.”

May combine two claims.

A model's behavior can shift as applicant pools change, so a system acceptable at launch can develop a disparate impact over time; without ongoing outcome monitoring, no one detects the change.

“A model's behavior can shift as applicant pools change, so a system acceptable at launch can develop a disparate impact over time; without ongoing outcome monitoring, no one detects the change.”

• AssumptionRests on an assumption: A model's behavior can shift as applicant pools change, so a system acceptable at launch can develop a disparate impact over time.

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

The same mis-framing that lets HR treat the tool as a neutral purchase lets IT treat it as neutral infrastructure: a system to be stood up and kept running, rather than one whose outputs carry legal and ethical weight.

“Yet the same mis-framing that lets HR treat the tool as a neutral purchase lets IT treat it as neutral infrastructure: a system to be stood up and kept running, rather than one whose outputs carry legal and ethical weight.”

• AssumptionRests on an assumption: Once a hiring tool is procured, it must be integrated, configured, connected to applicant data, and maintained, and these are IT responsibilities.