Digital HRM, Artificial Intelligence and the Future of Work

Introduction: HRM in a Digitally Connected Organisation

Human resource management has always depended on information: who works for the organisation, what capabilities employees possess, how people perform, what they need to learn, and why they leave. Digital technologies have transformed the speed, scale and form in which this information is collected and used. Human resource information systems (HRIS), cloud platforms, employee self-service, automation, people analytics and artificial intelligence (AI) now influence almost every major HR activity.

Digital HRM is not simply the replacement of paper forms with software. It is the redesign of HR processes through digital technologies, data and connected work systems. The objective is to improve administrative efficiency, employee experience, managerial decision-making and organisational capability. However, digitalisation also creates risks. Poor-quality data can produce poor decisions; algorithms can reproduce historical discrimination; excessive monitoring can damage trust; and cyberattacks can expose highly sensitive employee information.

The future of work is therefore not a purely technological question. It concerns the relationship between technology, jobs, skills, organisational design, employment regulation and human dignity. Effective HR professionals must be able to evaluate technology critically, not merely purchase or operate it.

Learning Objectives

By the end of this chapter, students should be able to:

1. Explain the purpose, architecture and applications of HR information systems.
2. Evaluate employee self-service, HR automation and people analytics platforms.
3. Assess applications of AI in recruitment, learning, performance management and workforce planning.
4. Identify algorithmic bias, privacy, cybersecurity and accountability risks.
5. Explain the importance of human oversight and responsible AI governance.
6. Analyse remote, hybrid, gig and skills-based work trends.
7. Recommend practical principles for implementing digital HR responsibly.

1. Digital HRM and the HR Technology Ecosystem

1.1 Meaning and scope

Digital HRM is the use of digital technologies, data and digital operating models to deliver, manage and improve HR work. It includes both transactional processes—such as payroll and leave administration—and strategic processes—such as workforce planning, capability development and organisational design.

The modern HR technology ecosystem commonly contains the following layers:

A useful distinction is between a system of record, a system of engagement and a system of intelligence. The system of record stores authoritative data, such as employment status. The system of engagement supports interactions, such as an employee requesting leave. The system of intelligence analyses data to support decisions, such as identifying skill shortages. A single platform may perform all three functions, but the concepts remain analytically distinct.

1.2 From administrative efficiency to strategic capability

Early HR systems focused on standardisation and cost reduction. Cloud-based systems have expanded access, integration and real-time reporting. Current digital HRM seeks to provide:

• consistent and auditable processes;
• faster service delivery;
• better employee access to information;
• improved workforce visibility;
• evidence-based decisions; and
• more personalised learning and career support.

Technology does not automatically create these outcomes. Benefits depend on process design, data quality, user adoption, managerial capability and governance. Automating a badly designed process may make errors faster and harder to detect.

2. HR Information Systems and Employee Self-Service

2.1 HRIS architecture and data quality

An HRIS generally receives data from recruitment, onboarding, payroll, time management, performance and learning systems. Integration reduces duplicate data entry but increases the importance of data governance. HR professionals must establish definitions for fields such as “active employee,” “headcount,” “turnover” and “completed training.” If departments use different definitions, apparently precise dashboards may give contradictory results.

Data quality has several dimensions:

• Accuracy: the record represents reality.
• Completeness: required fields are not missing.
• Consistency: the same concept is represented similarly across systems.
• Timeliness: data is updated when circumstances change.
• Validity: values conform to established rules.
• Uniqueness: duplicate records are controlled.

Data governance assigns responsibility for data ownership, access, retention, correction and audit. HR should not assume that an information system itself guarantees reliable information.

2.2 Employee and manager self-service

Employee self-service allows workers to access or update permitted information, request leave, select benefits, complete forms, view payslips and undertake learning. Manager self-service may support approvals, team dashboards, requisitions and performance conversations.

Self-service can improve speed and autonomy, but it may also transfer administrative work to employees without adequate support. Accessibility, digital literacy, language, disability inclusion and alternative channels must therefore be considered. A digital-first service should not become a digital-only service where that would exclude employees.

2.3 Automation and process redesign

HR automation may be rule-based or AI-enabled. Rule-based automation follows explicit conditions: for example, routing a leave request to the appropriate manager. AI-enabled systems infer patterns, classify text, recommend actions or generate content. The two should not be treated as identical.

Before automating a process, HR should ask:

1. Is the process necessary and legally defensible?
2. Is the desired outcome clearly defined?
3. Are inputs reliable and sufficiently structured?
4. What exceptions require human judgement?
5. How will errors be detected and corrected?
6. Who is accountable for the result?

A process map should identify the decision points, data inputs, system actions, human actions and escalation routes. Automation is most suitable for repetitive, high-volume and well-defined activities. It is less suitable where context, empathy, negotiation or moral judgement is central.

3. People Analytics: From Data to Decision

People analytics is the systematic use of workforce data, analytical methods and organisational knowledge to understand and improve people-related decisions. It is broader than producing HR reports. Reporting describes what happened; diagnostic analysis investigates why it happened; predictive analysis estimates what may happen; prescriptive analysis compares possible actions.

3.1 A disciplined analytics cycle

A credible analytics project follows a cycle:

1. Define a decision or business problem.
2. Formulate a measurable question.
3. Identify relevant data and its limitations.
4. Clean, link and analyse the data.
5. Interpret results in organisational context.
6. Test an intervention or decision.
7. Monitor outcomes, fairness and unintended effects.

The starting point should be a decision, not the mere availability of data. For example, “Which employees are likely to leave?” is less useful than “Which retention interventions should be offered to employees in critical roles, and under what conditions?”

3.2 Common HR metrics

A basic turnover rate can be expressed as:

The average headcount is often estimated as:

These measures assume that the period, population and definition of separation are clearly specified. Annualised and monthly rates should not be compared without considering seasonality and the method of annualisation.

For example, if an organisation records 30 separations during a quarter, with opening headcount of 480 and closing headcount of 520:

Average headcount = (480 + 520) ÷ 2 = 500

Turnover rate = (30 ÷ 500) × 100 = 6%

This is a quarterly rate. It should not automatically be described as an annual turnover rate.

Other useful measures include time to fill, time to productivity, absence rate, cost per hire, internal mobility, learning completion and employee experience. Every metric requires a definition, a population, a time period and an interpretation. A high learning completion rate, for instance, does not prove that learning improved performance.

3.3 Predictive analytics and responsible interpretation

Predictive models identify statistical associations between variables and an outcome. They do not establish causation automatically. A model may predict resignation accurately while offering little guidance about which intervention would prevent it. It may also use variables that act as proxies for protected characteristics, such as postcode, career interruptions or educational institution.

HR professionals should examine model performance across relevant groups, document assumptions, test for drift over time and avoid treating a probability as a certainty. Employees should not be labelled permanently because of a prediction.

4. Artificial Intelligence Across the HR Life Cycle

AI refers to computational systems that perform tasks associated with human cognitive activities, including classification, prediction, language processing, recommendation and content generation. Machine learning systems learn patterns from data; generative AI produces new text, images, code or other content in response to prompts.

4.1 Recruitment and selection

AI may support recruitment by drafting job descriptions, searching databases, matching candidates to role requirements, scheduling interviews, answering routine questions and assisting with structured assessment. These applications can reduce administrative burden, but selection is a high-risk domain because decisions affect access to employment.

Risks include:

• learning historical preferences that reflect discrimination;
• penalising non-traditional career paths;
• using irrelevant proxies for protected characteristics;
• excluding candidates with disabilities through inaccessible assessments;
• generating persuasive but inaccurate candidate summaries; and
• creating false confidence through apparently objective scores.

A responsible recruitment process begins with validated job-related criteria. Human reviewers should be trained to challenge recommendations rather than merely approve them. Candidates should receive appropriate information about automated assessment and a route for requesting accommodation or review, subject to applicable law.

4.2 Learning and development

AI can recommend learning based on role requirements, demonstrated skills, career interests and learning history. It can provide adaptive practice, translate material, generate simulations and answer routine questions. The danger is over-personalisation: employees may be restricted to predicted roles rather than encouraged to explore unfamiliar opportunities. Recommendations should therefore support, not replace, employee agency and manager discussion.

4.3 Performance and employee experience

Text analytics may identify themes in surveys or open comments. Virtual assistants can answer policy questions. However, sentiment analysis may misinterpret humour, cultural expression, multilingual communication or disability-related language. Monitoring communications can also create a climate of surveillance. The purpose, proportionality and access rules for such systems must be explicit.

4.4 Workforce planning

AI can combine workforce data with business forecasts to estimate future labour or skill requirements. A simple workforce gap is represented as:

Here, demand is the number or capability of workers required at a specified future time, and internal supply is the expected available capacity after considering exits, movement, development and productivity. The expression is a planning identity, not a prediction formula. It is useful only when demand and supply are measured on comparable bases—for example, both in full-time-equivalent roles or both in defined skill capacity.

If forecast demand is 1,200 full-time-equivalent roles and forecast internal supply is 1,050, the gap is 150 full-time-equivalent roles. The response could include hiring, reskilling, redesigning work, using technology or adjusting the business plan. The number alone does not determine the best intervention.

5. Algorithmic Bias, Fairness and Human Oversight

5.1 How bias enters digital HR

Bias can enter through historical data, problem definition, sampling, measurement, feature selection, model design, implementation and human use. A model trained on previous hiring decisions may reproduce the organisation’s past preferences. A performance model may mistake visibility for contribution in hybrid work. Even a technically accurate model can produce unfair outcomes if it is applied to an inappropriate decision.

Bias is not limited to intentional prejudice. It can arise from structural inequality, unequal access to opportunities, missing data and apparently neutral rules that have unequal effects.

5.2 Selection-rate analysis

One basic fairness diagnostic compares selection rates between groups:

A selection-rate ratio for a focal group compared with a reference group is:

For example, if 24 of 100 applicants in Group A are selected and 40 of 100 applicants in Group B are selected, their rates are 24% and 40%. The ratio for Group A relative to Group B is 0.60. In some jurisdictions, a ratio below 0.80 is used as a screening indication of possible adverse impact, commonly called the four-fifths rule. It is not a universal legal test, does not prove discrimination, and should not be used without context, statistical analysis, job-related validation and legal advice.

Fairness also involves false-positive and false-negative errors, calibration, accessibility, explainability and the effects of decisions. Improving one fairness measure may worsen another. Consequently, HR should not claim that an algorithm is simply “fair” without specifying the fairness objective and its limitations.

5.3 Human oversight as a control system

Human oversight is meaningful only when the reviewer has authority, competence, time and information to question the system. A person who can merely click “approve” is not effective oversight. Good controls include:

• documented decision roles and escalation routes;
• pre-deployment validation;
• testing across relevant demographic and occupational groups;
• explanation and review procedures;
• audit logs and version control;
• periodic monitoring for drift and disparate impact;
• correction, appeal and redress mechanisms; and
• suspension or retirement criteria.

Human involvement does not automatically remove bias. Humans may defer to automation, a phenomenon known as automation bias. Oversight must therefore be designed as a genuine challenge process.

6. Privacy, Cybersecurity and Governance

Employee data may include identity information, compensation, health information, biometric data, performance records, location, communications and inferred characteristics. HR has a heightened duty to use such data lawfully, fairly and proportionately.

6.1 Privacy principles

Responsible HR data practices generally require:

• a clear and legitimate purpose;
• data minimisation;
• transparency about collection and use;
• accuracy and correction rights;
• appropriate retention limits;
• access controls;
• secure processing; and
• accountability for vendors and internal users.

The precise legal requirements vary by jurisdiction. Organisations must therefore consult applicable privacy, employment, equality and AI regulation rather than relying on generic principles alone.

A useful governance question is whether the data is necessary for the stated decision. The fact that technology can collect keystrokes, location or facial signals does not establish that it should do so. Continuous monitoring may undermine autonomy, encourage gaming and create inaccurate interpretations of work.

6.2 Cybersecurity

HR systems are attractive targets because they contain valuable personal and financial information. Essential controls include multi-factor authentication, least-privilege access, encryption, secure configuration, patching, backups, incident response, staff training and supplier due diligence. Access should be role-based and regularly reviewed. Sensitive data should not be copied into unapproved generative AI tools.

Cybersecurity is also a people issue. Phishing awareness, clear reporting channels and a non-punitive approach to early incident reporting improve resilience. A technically secure system can still fail if employees are pressured to bypass controls.

6.3 Governance and accountability

A governance framework should identify:

Procurement contracts should address data ownership, confidentiality, security, subcontracting, model changes, audit rights, service continuity and deletion or return of data. HR should involve legal, information security, data protection, procurement, employee representatives and affected users where appropriate.

7. The Future of Work

7.1 Remote and hybrid work

Remote work separates work from a fixed organisational location; hybrid work combines remote and on-site work. These arrangements can improve flexibility, widen access to talent and reduce commuting. They can also produce isolation, unequal visibility, coordination problems, home-working burdens and blurred work–life boundaries.

Effective hybrid work is designed around work tasks rather than arbitrary attendance. Organisations should clarify which activities require co-location, establish communication norms, measure outcomes rather than presence, support ergonomics and accessibility, and ensure that remote workers receive equal access to information, development and promotion. Managers need capability in goal setting, coaching and inclusive communication.

7.2 Gig and platform work

Gig work involves short-term, task-based or project-based work, often arranged through digital platforms. It can provide flexibility and rapid access to specialised labour, but may involve income volatility, limited bargaining power, uncertain employment status and reduced access to benefits. HR professionals must distinguish legal classification from managerial preference and consider ethical responsibilities beyond minimum compliance.

Organisations using contractors should address quality, safety, confidentiality, data access, intellectual property, payment reliability and transition risks. A fragmented workforce can also weaken organisational learning and culture if knowledge is not deliberately integrated.

7.3 Skills-based organisations

A skills-based approach organises talent decisions around capabilities rather than relying solely on job titles, degrees or traditional career paths. Skills taxonomies and skills ontologies can support internal mobility, targeted learning and workforce planning. However, skills are difficult to define and measure. A digital profile may capture certificates while missing tacit knowledge, judgement and interpersonal contribution.

Skills-based HR should therefore combine structured data with manager and employee dialogue. It should avoid treating people as collections of marketable attributes. Fair access to learning and transparent criteria for opportunity allocation are essential.

7.4 Augmentation rather than simple substitution

Technology may automate tasks without eliminating entire occupations. It may also create new tasks, change skill requirements and increase the value of human abilities such as judgement, empathy, creativity, ethical reasoning and relationship-building. Workforce planning should analyse tasks and workflows, not just job titles. Reskilling, job redesign and transition support should be considered before redundancy where feasible.

8. Implementing Digital HR Responsibly

A practical implementation sequence is:

1. Diagnose the problem. Establish the user, decision, pain point and desired outcome.
2. Map the process. Identify data flows, exceptions, affected groups and accountability.
3. Assess alternatives. Compare process improvement, rule-based automation, AI and human-only approaches.
4. Evaluate risk. Consider privacy, security, bias, accessibility, employment impact and legal requirements.
5. Pilot carefully. Use a limited, representative setting with predefined success and stop criteria.
6. Validate outcomes. Examine effectiveness, user experience, error rates and group impacts.
7. Train and communicate. Explain purposes, limits, responsibilities and appeal routes.
8. Deploy with controls. Maintain logs, permissions, documentation and support.
9. Monitor continuously. Review drift, complaints, fairness, security incidents and unintended consequences.
10. Improve or retire. A system should be changed or withdrawn when it no longer meets its purpose responsibly.

A balanced evaluation should include efficiency, quality, employee experience, equity, compliance and strategic value. Cost savings alone are an insufficient measure of successful digital HR.

Applied Case: Illustrative Scenario—Northstar Services

Background

Northstar Services, a fictional customer-support organisation with 2,400 employees, experienced high turnover among entry-level advisers. It purchased an AI recruitment tool that ranked applicants using application histories, online tests and previous hiring data. It also introduced a dashboard predicting which current employees were likely to resign.

Problem

After six months, managers noticed that applicants from two local colleges were rarely shortlisted. Employees also complained that the resignation dashboard labelled them “high risk” without explanation. An internal review found that historical hiring data favoured applicants with uninterrupted employment histories. The turnover model used overtime, schedule changes and absence records, but did not account for managers’ unequal allocation of shifts.

Analysis

The recruitment model was not necessarily intentionally discriminatory, but its data and features reflected historical practices. The college variable may have operated as a proxy for socioeconomic background. The turnover model risked treating consequences of poor scheduling as employee risk. Both systems created accountability problems because managers relied on scores without understanding their limitations.

The HR director formed a review group including recruitment, operations, data protection, information security, employee representatives and an accessibility specialist. The group paused automated ranking, preserved records for audit and informed affected managers that the scores were advisory and under review.

Approach and solution

Northstar replaced college prestige and uninterrupted employment history with validated, job-related criteria. It introduced structured work-sample assessments, accessibility adjustments and trained human review. The turnover dashboard was redesigned as an aggregate team-level diagnostic rather than an individual label. Managers were shown patterns in scheduling, workload and development access, and were required to discuss support options rather than predict employee behaviour.

The organisation established quarterly fairness and security reviews, a candidate inquiry process, role-based access and a policy prohibiting use of dashboard scores in disciplinary decisions. It also measured whether interventions improved retention without creating unequal treatment.

Outcome

In the first review cycle, shortlisting became more diverse and the organisation identified scheduling instability as a stronger actionable issue than individual resignation risk. The technology remained useful for administration and aggregate planning, but high-impact individual decisions were returned to structured human processes.

Learning points

1. Historical data is not automatically an appropriate basis for future decisions.
2. A prediction may identify correlation without identifying a fair or effective intervention.
3. Aggregate analytics can sometimes provide organisational insight with less privacy risk than individual surveillance.
4. Human oversight requires authority, training and a genuine ability to challenge the system.
5. Responsible implementation includes the option to pause or withdraw technology.

Challenges and Limitations

Digital HR can increase standardisation while reducing discretion that is sometimes necessary for exceptional circumstances. Integrated systems can make errors propagate across multiple processes. Analytics can encourage metric fixation, where measurable activity replaces meaningful contribution. AI-generated content may be inaccurate, overconfident or based on undisclosed sources. Vendor claims may exceed independent evidence, and proprietary systems may limit auditability.

There are also distributional questions. Who benefits from flexibility? Who bears the cost of monitoring? Who receives high-quality learning recommendations? Which workers are visible in organisational data, and which are treated as external or temporary? A responsible HR professional examines these questions rather than assuming that innovation is neutral.

Chapter Summary

Digital HRM uses connected systems, data and digital operating models to improve HR administration, employee experience and strategic decision-making. HRIS platforms provide systems of record, engagement and intelligence, but their value depends on data quality, process design and governance. Employee self-service increases access and speed but must remain inclusive and supported.

People analytics progresses from descriptive reporting to diagnostic, predictive and prescriptive analysis. Metrics require precise definitions and should inform decisions rather than replace judgement. AI can support recruitment, learning, employee services and workforce planning, yet historical data, proxies and poor problem definitions can create algorithmic bias.

Responsible digital HR requires privacy protection, cybersecurity, transparency, accessibility, fairness testing, human oversight, auditability and routes for correction and appeal. The four-fifths rule may screen for possible adverse impact but is not a complete or universal fairness test. Remote, hybrid, gig and skills-based work create opportunities for flexibility and capability development while also raising concerns about equity, security, classification and surveillance.

The future of work is best understood as work redesign, not technology substitution alone. HR professionals must combine digital literacy with employment knowledge, ethical reasoning, critical analysis and human-centred leadership.

Key Terms

Artificial intelligence: Computational systems that perform tasks associated with human cognitive activities.

Algorithmic bias: Systematic and unfair differences in outcomes produced or amplified by an algorithmic system.

Cloud HRIS: An HR information system delivered through remotely hosted computing infrastructure.

Data governance: Policies, roles and controls governing the quality, access, use, retention and protection of data.

Employee self-service: Digital access through which employees complete or manage permitted HR activities themselves.

Four-fifths rule: A screening guideline that may indicate adverse impact when one group’s selection rate is below 80% of another group’s rate.

Gig work: Short-term, task-based or project-based work, often arranged through a platform.

Human oversight: Meaningful human authority and competence to review, challenge and correct an automated system.

Hybrid work: A work arrangement combining remote and on-site work.

HR information system: A system that stores, processes and supports the use of human resource information.

Machine learning: Computational methods that learn patterns from data to classify, predict or recommend.

People analytics: Systematic analysis of workforce data to inform people-related decisions.

Predictive analytics: Analysis that estimates the likelihood of future outcomes.

Privacy by design: Incorporating privacy safeguards into systems and processes from the beginning.

Selection rate: The proportion of people considered who are selected.

Skills-based organisation: An organisation that structures talent decisions around capabilities rather than only jobs or credentials.

System of record: The authoritative source for defined organisational data.

System of engagement: A platform that supports interactions between users and the organisation.

Workforce planning: The process of aligning future workforce demand, supply, skills and interventions.

Review Questions

Short Answer Questions

1. What is digital HRM, and how does it differ from simply computerising paperwork?
2. Distinguish between a system of record, a system of engagement and a system of intelligence.
3. Identify three dimensions of HR data quality.
4. Why does human oversight require more than a person approving an automated recommendation?
5. State two cybersecurity controls relevant to HR systems.

Long Answer Questions

1. Explain the benefits and limitations of HR information systems and employee self-service.
2. Describe the people analytics cycle and explain why a decision problem should precede data collection.
3. Evaluate the uses and risks of AI in recruitment and learning and development.
4. Discuss how privacy, cybersecurity and algorithmic fairness should influence digital HR governance.
5. Compare remote, hybrid, gig and skills-based work in terms of opportunities and HR risks.

Analytical / Application Questions

1. An organisation has 650 employees at the beginning of a year, 750 at the end, and 70 separations during the year. Calculate the turnover rate using average headcount and explain one limitation of the result.
2. A recruitment model selects 18 of 90 applicants in Group A and 30 of 100 applicants in Group B. Calculate both selection rates and the selection-rate ratio for Group A relative to Group B. Interpret the result cautiously.
3. Design a governance plan for an AI tool that recommends employees for promotion. Specify the data checks, fairness tests, privacy safeguards, human decision rights and monitoring arrangements you would require.

Multiple Choice Questions with Answers

Question 1

Which HR technology is primarily a system of record?

A. A database maintaining authoritative employee and job information

B. A chatbot answering routine policy questions

C. A collaboration platform for informal discussion

D. A predictive model estimating resignation risk

Correct Answer: A. A database maintaining authoritative employee and job information

Question 2

What is the main purpose of people analytics?

A. To eliminate all managerial judgement

B. To use workforce data systematically to inform decisions

C. To collect as much employee data as possible

D. To replace employment law with statistical rules

Correct Answer: B. To use workforce data systematically to inform decisions

Question 3

An organisation has 500 employees at the beginning of a period and 600 at the end. It records 33 separations. What is the turnover rate using average headcount?

A. 5.5%

B. 6.0%

C. 6.6%

D. 11.0%

Correct Answer: B. 6.0%

Question 4

Which statement best describes a risk of using historical hiring data to train an AI recruitment tool?

A. Historical data always improves fairness

B. The system may reproduce previous discriminatory preferences

C. The system cannot process structured data

D. The system eliminates the need for job-related criteria

Correct Answer: B. The system may reproduce previous discriminatory preferences

Question 5

A selection-rate ratio below 0.80 under the four-fifths rule should be interpreted as:

A. Conclusive proof of unlawful discrimination

B. Evidence that the model is accurate

C. A possible adverse-impact warning requiring further investigation

D. Proof that the smaller group should receive automatic selection

Correct Answer: C. A possible adverse-impact warning requiring further investigation

Question 6

Which practice best demonstrates meaningful human oversight?

A. A manager approves every recommendation without seeing the evidence

B. A reviewer can investigate, challenge and correct system recommendations

C. An organisation hides the system’s purpose from employees

D. A vendor alone decides when the system should be used

Correct Answer: B. A reviewer can investigate, challenge and correct system recommendations

Question 7

Which is an example of data minimisation?

A. Collecting every available employee signal in case it becomes useful

B. Retaining all employee data permanently

C. Collecting only information necessary for a clearly defined purpose

D. Sharing HR data widely to improve transparency

Correct Answer: C. Collecting only information necessary for a clearly defined purpose

Question 8

What is a major risk of measuring hybrid-work performance mainly through online presence?

A. It may confuse visibility or activity with meaningful contribution

B. It guarantees equal access to promotion

C. It prevents all cybersecurity incidents

D. It measures outcomes more accurately than any other method

Correct Answer: A. It may confuse visibility or activity with meaningful contribution

Question 9

If forecast workforce demand is 1,000 full-time-equivalent roles and forecast internal supply is 920, what is the workforce gap?

A. 80 surplus roles

B. 80 full-time-equivalent roles

C. 920 full-time-equivalent roles

D. 1,920 full-time-equivalent roles

Correct Answer: B. 80 full-time-equivalent roles

Question 10

Which implementation sequence is most responsible for a high-impact AI HR system?

A. Purchase first, define the purpose later

B. Deploy immediately and investigate complaints only if they occur

C. Define the problem, assess risks, pilot, validate, monitor and improve

D. Transfer all accountability to the software vendor

Correct Answer: C. Define the problem, assess risks, pilot, validate, monitor and improve