# AI hiring procurement evidence A matching score, an assessment result and a payroll workflow do different jobs. This register links public evidence to the questions a buyer still needs answered. It does not rank vendors, certify systems or report hands-on tests. Author: Gene Dai Published: 2026-09-08 Updated: 2026-09-08 Canonical: https://digidai.github.io/research/ai-hiring/ ## Method and limits - Selection: six vendors discussed in this site's hiring and HR coverage, spanning assessment, application review, candidate matching and HCM. This is a curated starting set, not a representative market sample. - Each record describes one document and one procurement question. Checked means the available public text was reviewed on that date; excerpt-only access is stated in the record scope. It does not mean a full-document audit, customer deployment or contract was verified. An undated source remains undated. - A vendor-published independent audit summary is labeled by both its publisher and its scope. It is not evidence that every deployment is unbiased, legally compliant or effective at predicting performance. - Ask for the exact product version, module, language, geography, data flow and decision authority. Validate the configured workflow with authorized, representative data before making employment decisions. This resource is not legal advice. - Editorial disclosure: Gene Dai is a co-founder of Metix AI, a recruitment technology company. This register has no paid placement or affiliate links; the commercial relationship still matters when interpreting the analysis. ## Using this register ### Before a demo Pick the workflow you actually buy. Send the matching record's next-check question to the vendor and ask for the underlying document, not a yes/no slide. ### During a pilot Record human review, overrides, failures, completion rates and subgroup sample sizes alongside time saved. Agree the acceptance test and stop conditions before seeing the results. ### Before a renewal Check whether the model, criteria, permissions, subprocessors or commercial unit changed. An earlier audit and an earlier quote may no longer describe the same service. ## Evidence records ### SHL: Does an AI assessment still need job-related validation? Record: [shl-assessment-validity](https://digidai.github.io/research/ai-hiring/#shl-assessment-validity) SHL's AI principles require reliability, validity and job relevance for AI assessments, as for other assessments. The document also treats explainability as a design requirement. - Evidence type: Vendor documentation - Source: [SHL: Core Principles for the Ethical and Effective Use of AI to Assess Talent](https://www.shl.com/assets/documents/how-shl-is-using-ai-and-machine-learning-white-paper.pdf) - Source date: Not stated on source - Source checked: 2026-09-08 - Scope: Publicly indexed excerpt of vendor AI-assessment principles; direct document retrieval failed during review. Not a technical manual for a named test or employer. - Limitation: Principles alone do not establish predictive validity, accessibility or subgroup performance for your assessment, language and applicant population. - Next check: Request the technical manual, validation cohort, norm group, accommodation process and monitoring plan for the exact instrument and scoring version. ### HireVue: Which input and output does the assessment actually use? Record: [hirevue-explainability](https://digidai.github.io/research/ai-hiring/#hirevue-explainability) HireVue publishes an AI explainability statement describing its assessment approach. Treat that statement as a starting point for tracing candidate inputs, scoring and human decisions. - Evidence type: Vendor documentation - Source: [HireVue 2024 AI Explainability Statement](https://www.hirevue.com/wp-content/uploads/2024/09/HV_2024_AI-Explainability-Statement.pdf) - Source date: 2024 - Source checked: 2026-09-08 - Scope: Indexed excerpt and publisher description of the named 2024 statement; full-PDF retrieval exceeded the review tool's size limit. It cannot automatically describe later products or a customer's current configuration. - Limitation: A public explanation does not verify the current model, enabled modalities, score use or retention settings in a prospective deployment. - Next check: Ask for the current statement, exact input modalities, model version, candidate notice, deletion controls and how an assessor can contest or override an output. ### Ashby: Who defines the criteria used to review applications? Record: [ashby-application-review](https://digidai.github.io/research/ai-hiring/#ashby-application-review) Ashby documents job-specific resume criteria for AI-assisted application review, with organization-admin enablement and role-level permissions for editing criteria. - Evidence type: Vendor documentation - Source: [Ashby: AI-Assisted Application Review](https://docs.ashbyhq.com/ai-assisted-application-review) - Source date: Not stated on source - Source checked: 2026-09-08 - Scope: Documented configuration of application review, not observed screening accuracy or a guarantee of plan entitlement. - Limitation: A resume can omit a skill the candidate has. Criteria matching is not a validated measure of future job performance. - Next check: Test ambiguous and incomplete resumes; identify who can change criteria, how changes are logged and whether a reviewer sees the underlying evidence before rejecting anyone. ### Greenhouse: Does the audit cover the feature you plan to enable? Record: [greenhouse-ai-audit-claims](https://digidai.github.io/research/ai-hiring/#greenhouse-ai-audit-claims) Greenhouse's AI principles page says Talent Matching and AI Interviewer undergo independent monthly bias audits by Warden AI. This record captures the vendor's claim, not a review of those audit reports. - Evidence type: Vendor documentation - Source: [Greenhouse: Responsible AI in Recruitment for Hiring Teams](https://www.greenhouse.com/ai-principles) - Source date: Not stated on source - Source checked: 2026-09-08 - Scope: Named features and vendor-stated audit cadence on a public principles page. - Limitation: An audit cadence does not specify the dataset, tested model version, selection thresholds, exceptions or applicability to a customer. - Next check: Obtain the latest feature-specific report, auditor identity, test dates, sample sizes, excluded groups and documented limitations. Compare them with the deployment you are buying. ### Workday / HiredScore: Can one employer's audit stand in for yours? Record: [workday-spotlight-audit-scope](https://digidai.github.io/research/ai-hiring/#workday-spotlight-audit-scope) Workday publishes a Secretariat analysis of its own Spotlight deployment. Testing completed March 20, 2026; the stated applicant period is September 1, 2025 through February 28, 2026. - Evidence type: Vendor-published audit summary - Source: [Workday: Responsible AI and Bias Mitigation for HiredScore Spotlight](https://www.workday.com/en-us/legal/responsible-ai-and-bias-mitigation.html) - Source date: Not stated on source - Source checked: 2026-09-08 - Scope: U.S. residents applying for greater-New-York-City roles in Workday's own hiring; selected high-volume job profiles and demographic-data exclusions are described in the source. - Limitation: Workday explicitly limits the analysis to its implementation and says the page does not satisfy a customer-specific legal obligation. It is not a universal fairness certificate. - Next check: Compare your job mix, thresholds, candidate pool and human review with the reported scope; request evidence for the currently deployed model and your own use case. ### UKG: Which contracted module performs the promised workflow? Record: [ukg-contract-scope](https://digidai.github.io/research/ai-hiring/#ukg-contract-scope) UKG publishes separate service descriptions for Pro HCM, Pro Workforce Management, Ready and AI-related services. A product-family name is not a complete statement of the purchased scope. - Evidence type: Vendor documentation - Source: [UKG: Services Descriptions](https://www.ukg.com/legal/services-descriptions) - Source date: Not stated on source - Source checked: 2026-09-08 - Scope: Public service-description directory, not an executed order form, statement of work or customer entitlement. - Limitation: A feature appearing in the portfolio does not prove it is included, configured, supported in a country or connected to the employer's ATS. - Next check: Map each promised action to a module, order-form line, implementation owner, data integration, country scope and acceptance test. ## Related reading - [SHL assessments: candidate help and buyer checks](https://digidai.github.io/2025/07/17/shl-deep-analysis/) - [UKG Pro: scope, pricing questions and implementation checks](https://digidai.github.io/2025/08/09/ukg-pro-comprehensive-enterprise-analysis/) - [AI and employment evidence](https://digidai.github.io/research/ai-employment/) ## Update log - 2026-09-08: Initial six-vendor register. Public documentation and a deployment-specific audit summary are separated from product testing; missing contract and pilot evidence remains explicit. ## Formats and corrections [JSON](https://digidai.github.io/research/ai-hiring/data.json) | [CSV](https://digidai.github.io/research/ai-hiring/data.csv) | [Review worksheet](https://digidai.github.io/research/ai-hiring/review-template.csv) | [Research updates RSS](https://digidai.github.io/research/feed.xml) | [Corrections](https://digidai.github.io/contact/) The review worksheet contains unanswered questions and blank fields for your evidence. It is not a completed audit or test result. These are editorial summaries, not a license to redistribute the linked publications. Cite original sources and preserve limitations.