AI-Powered Human Resource Management in the Public Sector

• Changing nature of public administration
• HR’s role in building a capable developmental state
• Workforce pressures facing Southern African governments
• Skills shortages and competition for digital talent
• Ageing workforces and succession risks
• Fragmented HR information systems
• Slow recruitment and appointment processes
• Public-sector productivity challenges
• Meaning of AI-powered HR management
• Introduction to: Artificial intelligence; Machine learning; Generative AI; Natural-language processing; Predictive analytics; Intelligent automation; Robotic process automation; AI-enabled chatbots
• Difference between AI, automation, analytics and conventional software
• Descriptive, diagnostic, predictive and prescriptive analytics

• Data as the foundation of AI-powered HR
• Public-sector HR data architecture
• Employee master data
• Establishment and position data
• Skills and competency data
• Recruitment and applicant data
• Learning and performance data
• Payroll, benefits, attendance and leave data
• Dimensions of data quality: Accuracy; Completeness; Consistency; Timeliness; Validity; Uniqueness
• Data ownership and stewardship
• Data standards and interoperability
• Master data management
• Integrating fragmented HR systems
• Data lineage
• Data retention and disposal
• Data-quality dashboards
• Linking workforce planning to government priorities
• Analysing current workforce supply
• Forecasting future workforce demand
• Identifying critical and scarce skills
• Skills-based workforce planning
• Workforce segmentation
• Predictive retirement analysis
• Succession planning
• Workforce cost analysis
• Scenario modelling
• Identifying redeployment opportunities
• Internal mobility and talent marketplaces
• AI-assisted skills matching
• Risks of incomplete or outdated employee profiles
• Human validation of workforce recommendations

Recruitment and selection

• Public-sector merit and fairness principles
• AI-supported job descriptions and vacancy drafting
• Skills-based job profiles
• Automating vacancy administration
• Applicant communication and interview scheduling
• CV and application analysis
• Candidate matching
• Digital assessments
• Recruitment chatbots
• Verification and fraud risks
• Reducing recruitment turnaround time
• Bias in training data
• Proxy discrimination
• Accessibility and digital exclusion
• Explainability of screening recommendations
• Human review and appeal mechanisms
• Documentation and audit trails

Onboarding and employee self-service

• Automating appointment documentation
• Digital induction
• Personalised onboarding pathways
• AI-based policy and procedure assistants
• HR chatbots
• Leave, payroll and benefit enquiries
• Multilingual employee support
• Accessibility and inclusive design
• Escalating complex enquiries to HR officials
• Monitoring chatbot accuracy
• Protecting employee information

Learning and reskilling

• Identifying public-service skills gaps
• Digital and AI literacy
• AI-supported training-needs assessment
• Personalised learning pathways
• Intelligent content recommendations
• Generative AI for training materials
• Microlearning
• Simulations and scenario-based learning
• Digital coaching tools
• Leadership development
• Communities of practice
• Measuring learning application
• Validating AI-generated training content
• Protecting intellectual property
• Reskilling employees affected by automation

Employee engagement and experience

• Employee experience in public institutions
• Continuous employee listening
• Digital pulse surveys
• Natural-language analysis of employee feedback
• Engagement and morale indicators
• Analysing change readiness
• Workforce well-being indicators
• Distinguishing organisational insight from employee surveillance
• Cultural and linguistic limitations of sentiment analysis
• Anonymous and confidential feedback
• Representation of regional, remote and frontline employees
• Validating findings through direct engagement
• Converting employee insights into management action

Generative AI for HR productivity

• Drafting HR correspondence
• Summarising policies and reports
• Developing interview questions
• Preparing training materials
• Creating frequently asked questions
• Producing meeting summaries
• Comparing policies
• Supporting dashboard development
• Effective prompt design for HR professionals
• Verifying AI-generated information
• Confidentiality risks
• Hallucinations and inaccurate outputs
• Approved versus unauthorised AI tools
• Maintaining records of significant AI use

Responsible AI principles

• Human dignity
• Fairness and non-discrimination
• Transparency
• Explainability
• Accountability
• Privacy
• Security
• Reliability
• Accessibility
• Inclusion
• Proportionality
• Human oversight
• Contestability and appeal
• Identification of high-risk and prohibited uses
• Balancing innovation with public trust

Privacy and cybersecurity

• Sensitivity of employee information
• Personal and special-category data
• Lawful processing
• Purpose limitation
• Data minimisation
• Informed employee communication
• Role-based access
• Identity and access management
• Encryption
• Cloud and data-transfer risks
• Third-party data processing
• Cybersecurity incidents and breach response
• Data retention and deletion
• Risks of using public generative AI tools
• Cross-border data hosting
• Privacy impact assessments
• Employees’ rights to access and correct information

Labour relations and human impact

• Employee fears concerning AI
• Automation and job redesign
• Consultation with employees and organised labour
• Communicating what AI will and will not do
• Role changes and redeployment
• Reskilling and upskilling
• Protecting employee dignity
• Managing resistance
• Procedural and distributive fairness
• Collective bargaining implications
• AI-supported disciplinary processes
• Human accountability in employment decisions
• Preventing technology-driven exclusion
• Creating safe channels for employee concerns

AI procurement and vendor management

• Developing a clear AI business case
• Avoiding technology-driven procurement
• Defining functional and technical requirements
• Data and model ownership
• Explainability requirements
• Vendor access to employee information
• Cloud and hosting arrangements
• Integration with legacy systems
• Cybersecurity requirements
• Service-level agreements
• Bias and accuracy testing
• Vendor lock-in
• Intellectual-property provisions
• Audit rights
• Incident notification
• Model updates and change control
• Exit and data-return provisions
• Monitoring vendor performance

AI governance and accountability

• Executive and board oversight
• Role of central public-service authorities
• AI and Data Governance Committee
• HR, ICT, legal, ethics and risk responsibilities
• Data-owner and system-owner responsibilities
• Employee and labour representation
• AI use-case registers
• AI risk classification
• Approval processes
• Model documentation
• Human review requirements
• Incident management
• Continuous monitoring
• Assurance and reporting
• Annual policy review

Role of Internal Audit

• Internal Audit’s role under the Three Lines Model
• Maintaining independence and objectivity
• Advisory versus assurance responsibilities
• Incorporating AI risk into audit planning
• Auditing: Governance; Data quality; Privacy; Cybersecurity; Recruitment algorithms; Performance analytics; Payroll analytics; Vendor arrangements; Human oversight; Bias and fairness
• Testing model outputs
• Assessing explainability
• Reviewing access and change controls
• Conducting post-implementation reviews
• Continuous auditing and monitoring
• Reporting material risks to oversight bodies

Measuring value and performance

• Establishing baseline information
• Benefits-realisation management
• Measuring: Recruitment turnaround time; HR processing time; Employee query resolution; Data accuracy; Payroll exceptions; Skills coverage; Learning application; Employee satisfaction; System adoption; Cost efficiency
• Measuring fairness and inclusion
• Monitoring false positives and false negatives
• Tracking employee complaints
• Assessing unintended consequences
• Deciding whether to scale, modify or terminate a solution

• Leadership responsibilities in AI transformation
• Establishing the case for change
• Executive sponsorship
• Stakeholder mapping
• Change-impact assessment
• Communication and employee engagement
• Leadership and managerial digital literacy
• Manager coaching
• Resistance management
• Pilot-based implementation
• Employee support mechanisms
• Adoption and usage measurement
• Reinforcement of new practices
• Managing change fatigue
• Coordination across government departments
• Capturing and applying lessons learned
• Scaling successful AI solutions
• Sustaining new HR practices
• Developing an implementation roadmap
• Monitoring strategic risks and benefits