W3: Rodgers et al. (2023)

Question 1

Article’s purpose

Answer 1

Focuses on the integration of ethical considerations into algorithmic decision-making within HRM processes. It addresses the increasing use of AI in HRM and the necessity for a framework that ensures ethical accountability in these AI-driven processes. This study proposes grounding AI-driven HR in six ethical frameworks to create more accountable systems. It introduces a throughput model and provides tools for auditing AI decisions in pay equity, diversity initiatives, and performance management, offering HR professionals a blueprint for ethical AI implementation that protects both organisational goals and employee rights

Question 2

Current HRM AI implementations

Answer 2

Often prioritise efficiency over ethical considerations, particularly in gig economy applications where algorithmic management replaces human oversight

Question 3

Throughput Model

Answer 3

Integrates six ethical frameworks principles into AI design, enabling organisations to balance automation’s benefits with fairness. It offers insights from cognitive and social psychology into a descriptive model of how human constituents make decisions within organisations. In the first stage, both P and I influence J, then, in the second stage, P and J influence decision choice. It provides a broad conceptual framework for examining the interconnected processes that influence decision choices within organisations. It offers a framework for communication and understanding AI-driven HRM decisions

Question 4

“Environmental variables”

Answer 4

Internal environment, encompassing natural social, and economic factors. An organisation’s intent is shaped by its internal environment. The integration of AI technology, incorporating these into HRM algorithms, provides an opportunity for post-decision evaluation through root cause analysis (RCA)

Question 5

Solutionism

Answer 5

The failure to recognise that the optimal solution to a problem may not always involve technology

Question 6

The ripple effect

Answer 6

The failure to fully understand how the incorporation of technology into an existing social system alters the behaviours and embedded values of that system

Question 7

Formalism

Answer 7

The inability to account for overall connotation of social concepts, such as fairness, which are procedural, contextual, and contestable, and thus cannot be fully captured through mathematical formalisms

Question 8

Portability

Answer 8

The failure to comprehend that algorithmic solutions developed for one social context may be misleading, erroneous, or detrimental when applied to a different context

Question 9

Framing

Answer 9

The failure to model the complete system, including the social criteria, such as fairness, that will be enforced

Question 10

AI

Answer 10

A technology that seeks to simulate human reasoning in computers and other machines

Question 11

Algorithms

Answer 11

Sets of unambiguous specifications for performing tasks such as calculations, data processing, and automated reasoning. They are fundamental to AI

Question 12

Time pressure decisions

Answer 12

The cost of unhurried decisions is high (spedd being essential)

Question 13

Accuracy

Answer 13

The cost of wrong decision choices is minimised

Question 14

Allocation of resources

Answer 14

The data size is too large for manual analysis or traditional algorithms

Question 15

Decision accountability framework

Answer 15

Study indicates a need for introducing it whereby HRM practitioners have a pathway to consider and account for components of the organisational environment, employee engagement, and ethics when incorporating AI decision-making to assist in achieving organisational goals

Question 16

Anomaly detection

Answer 16

Identify items, events, or observations that do not conform to an expected pattern or other items in a pool of job applicants

Question 17

Background verification

Answer 17

Machine learning-powered predictive models can extract meaning and highlight issues based on structured and unstructured data points from applicants’ resumes

Question 18

Employee attrition

Answer 18

Find employees who are at high risk of attrition, enabling HR to proactively engage with and retain them

Question 19

Content personalisation

Answer 19

Provide a more personalised employee experience by using predictive analytics to recommend career paths, professional development programs, or optimise a workplace environment based on prior employee actions

Question 20

Deep learning

Answer 20

A branch of machine learning that trains a computer to learn from large amounts of data through neural network architecture. It is a more advanced form of machine learning that breaks down data into layers of abstraction

Question 21

Image and video recognition

Answer 21

Can classify candidates based on objective data and predict fraudulent behaviour using behavioural analytics. HRM practitioners need to incorporate ethical frameworks when using real-time AI psychological profiling systems that analyse non-verbal behaviour

Question 22

Speech recognition

Answer 22

Enables virtual assistants and speech analytics software for compliance, fraud detection, and communication review. However, the data collected may contain sensitive information, necessitating ethical guidelines for its collection, processing, and storage

Question 23

Chatbots

Answer 23

Utilising Natural Language Processing (NLP) are becoming crucial for automating HRM service delivery

Question 24

Recommendation engines

Answer 24

In digital learning, personalise learning pathways and provide managers with training suggestions. However, reliance on AI decision-making in performance management may lead to a replication of human-machine contact and a deferral of responsibility, potentially resulting in a negative employee experience

Question 25

Perception (P)

Answer 25

Shaped by experience, education, and biases. It is a process of individuals framing their problem-solving set or view of the world

Question 26

Information (I)

Answer 26

Continuously updates perception, similar to Bayesian learning, including the set of technical, managerial, economic, political, social, and environmental information available to a decision maker for problem-solving

Question 27

Judgement (J)

Answer 27

Weighs and analyses inputs, containing the process by which individuals implement and analyse incoming information and the influences from their perception

Question 28

Decision (D)

Answer 28

The final decision taken (or not taken)

Question 29

P <--> I relationship

Answer 29

Functions like a neural network, where perception adjusts based on new data, enabling AI systems to learn iteratively

Question 30

Preference-based (ethical egoism)

Answer 30

Decision maximise self-interest, ignoring external information. This theory posits that individuals should act in their own self-interest. In an AI-driven HRM context, this could translate to algorithms that prioritise decisions that benefit the organisation, potentially at the expense of individual employees. P -> D

Question 31

Algorithmic pathways

Answer 31

Six can be used by a decision maker. They influence a decision choice and reflect the problem statement in the introduction, whereby the modelling process may help arrest problems of transmitting and receiving HRM knowledge and information due to organisations seeking different and comparative ethical solutions to a problem.

Question 32

Rule-based (deontology)

Answer 32

Decisions follow moral rules (e.g. laws, policies). This rule-based theory emphasises adherence to moral duties and obligations. In AI-driven HRM, this would involve algorithms that follow pre-defined ethical, such as fairness and respect for employee rights. P -> J -> D

Question 33

Principles-based (utilitarianism)

Answer 33

Decisions maximise societal good using objective data. This principle-based theory advocates for actions that maximise overall happiness or well-being. AI-driven HRM systems based on it would aim to make decisions that benefit the greatest number of employees, even if some individuals are negatively affected. I -> J -> D

Question 34

Relativism-based

Answer 34

Ethics depend on situational/cultural context. This theory suggests that ethical standards are relative to individual cultures or societies. In AI-driven HRM, this could lead to algorithms that adapt to different cultural norms, potentially resulting in inconsistent ethical standards across different contexts. I -> P -> D

Question 35

Virtue ethics-based

Answer 35

Decisions reflect moral character (e.g. integrity, fairness). This theory focuses on developing good character traits. AI-driven HRM systems based on virtue ethics would prioritise decisions that reflect virtues such as fairness, honesty, and compassion. P -> I -> J -> D

Question 36

Ethics of care-based (stakeholders)

Answer 36

Decisions prioritise stakeholder well-being (e.g. empathy, equity). This stakeholder-centred theory emphasises the importance of relationships and caring for the well-being of others. In AI-driven HRM, this would involve algorithms that consider the impact of decisions on all stakeholders, including employees, employers, and the broader community, prioritising the needs of the most vulnerable. I -> P -> J -> D

Question 37

Type 1 errors

Answer 37

False positive. May fuel inefficiencies and increase transaction costs, which can cause inadequate algorithms, as depicted by an AI system

Question 38

Type 2 error

Answer 38

False negative. May engender inappropriate workforce individuals to receive opportunities

Question 39

Decision-making

Answer 39

Involves assessing action initiation and comparing expected versus actual results based on the decision maker's intent. Accountability for AI decisions is crucial, requiring understanding organisational hierarchies and the influence of organisational culture on ethical AI use

Question 40

AI opacity

Answer 40

Due to this, HRM needs to mediate human involvement levels for accountability. The interaction between people and organisations is key, and the TP model can analyse AI decision-making at the organisational level

Question 41

Human involvement

Answer 41

Depends on understanding objectives within the organisational hierarchy and external environment

Question 42

AI neural networks

Answer 42

Learn continuously

Question 43

AI algorithmic ethics framework

Answer 43

Fundamental to minimise bias, and human intervention combined with AI can foster unbiased HRM practices

Question 44

Root-Cause Analysis (RCA)

Answer 44

Help assess the level of human involvement and detachment from AI decisions. E.g., in cases where software is adopted under time pressure without adequate staff training, unanticipated negative outcomes can arise, underlining the need for HRM to evaluate decision pathways holistically

Question 45

Decision Dashboard

Answer 45

Ultimately enables a continuous, iterative process to refine AI integration in alignment with ethical considerations and organisational goals