Record STMT-WF6B345Soft law

AI in human resource management: The limits of empiricism

Report Institutional policyGlobaladopted

This working paper examines the integration of artificial intelligence into human resource management across four core functions: recruitment, compensation, scheduling, and performance management. It introduces a three-parameter framework assessing AI systems based on their objectives, data, and programming, arguing that reductive metrics and opaque algorithms pose risks to workers and firms. The authors contend that an uncritical reliance on empiricism overlooks the limits of capturing complex human behaviour through data.

Summary and content fields are drawn from the document text and are pending editorial review.

Who issued it

International Labour OrganizationIntergovernmental

Lead issuer

Legal forceSoft law

Under whose authorityInstitutional policy

Judged fromfull text

Who it governs

Researchers and academicsNational governmentCompaniesWorkers

This paper is for human resource professionals, researchers, policymakers, and employers seeking to understand the limits and risks of integrating artificial intelligence into workforce management.

Sectors

Labour and employment · human resource management

What the document contains

What's covered

This publication reviews how artificial intelligence is applied across four core human resource management functions: recruitment, compensation, compensation structure, scheduling work, and performance management. It analyses structural shortcomings such as poorly aligned objectives, biased datasets, and opaque programming, and outlines a framework for evaluating these technologies.

What's missing

The document does not provide binding legal regulations, enforcement mechanisms, or prescriptive compliance checklists for software developers.

Key takeaways

AI adoption in human resource management is growing rapidly despite a lack of comprehensive assessment regarding its usefulness.
AI systems in HR can be evaluated through a three-parameter framework: system objective, underlying data, and programming.
Recruitment, compensation, scheduling, and performance management face significant legal, ethical, and practical risks due to reductive metrics and biased training data.
An uncritical reliance on positivism and numeracy in HR has fostered unbridled optimism that overlooks the complexities of human behaviour.
Effective implementation of AI in the workplace requires participatory, multi-stakeholder design involving HR professionals and workers.

Outline

· Abstract
· About the authors
· Introduction
· 1 How did we get here? The rise of “people analytics” and managing workers through data
· 2 The workings of AI systems: Objective, data and programming
· 3 AI in HRM: Unbridled optimism
· 4 What is an HR manager to do with AI?
· Conclusion
· References
· Acknowledgements

Issued11 November 2025

Length17,073 words

FormatNot recorded

Scopeglobal

How it scores

Pending review

Each score is backed by a quote verified against the document text. Editorial review has not yet confirmed them.

Enforcement

0/5

Consequences and compliance mechanisms with real force

screening score · 2026-10-23

Impact Assessment

2/5

Required assessment of an AI system's risks/impacts before or during use

As a public institution, the Treasury Board completed an Algorithmic Impact Assessment to identify possible risks from using the system.

Screening, Interviewing and Selection

screening score · 2026-10-23

Topics

artificial_intelligence human_resource_management algorithmic_management workforce_analytics labour_standards