Planning the Workforce for AI-Driven Change: Roles, Skills, Leadership and Organisational Readiness
A brief opinion and personal view of the challenges and opportunities which lie ahead for health and social care leadership, management and staff.
8 July 2026·Mark Cantwell·12 min read
Executive summary
AI, AGI-like capabilities, interoperability, agentic automation, and connectivity are changing workforce planning, talent strategy, and organisational design. Certain roles are more exposed; others are likely to evolve. New challenges will emerge for people-management and organisational development capabilities, and so organisational capability itself will be forced to evolve to meet them.
“Technology changes the work, but people strategy determines whether the organisation gains from it.
— Incredible Health, 2026 Executive Report
Preparing for competence upgrades, new competency families and the impact of AI/AGI will require focussed planning for learning, coaching and training — matched by Organisation Design and Development (OD) initiatives that shape person-centred, learning-oriented values, policies and processes. Health and social care will be front and centre as highly regulated, systems-heavy external and internal processes evolve in response.
In this piece
- 1Technologies driving change
- 2Roles likely to disappear or evolve
- 3Competitive and talent implications
- 4People and OD implications
- 5Policies and readiness
- 6Assessments and actions needed to be made now
- 7Further reading
1. Technologies driving change
The main shift is from digitising tasks to orchestrating work across humans, software agents, embedded analytics, and connected systems. AI and big data are among the fastest-growing skills areas, while employers also report rising demand for networks, cybersecurity, and technology literacy. Interoperability matters because organisations increasingly need systems, data, and workflows that can connect across platforms, vendors, and functions without manual handoffs.
Agentic efficiency will especially affect knowledge work: scheduling, triage, drafting, analysis, service routing, and routine decision support can increasingly be delegated to AI systems or semi-autonomous agents. That does not remove the need for people — it changes the human role from doing the task to supervising exceptions, assuring quality, and handling ambiguity.
2. Roles likely to disappear or evolve
Roles most at risk are those built around routine, repeatable, rules-based, or transactional work — data entry, basic administration, payroll processing, routine bookkeeping, call-centre handling, standard customer service, cashiering, and some scheduling or front-office functions. These roles may not vanish completely, but the number of posts is likely to shrink sharply as self-service, automation, and AI support expand.
Roles likely to evolve include HR, finance, procurement, operations, customer service, middle management, and compliance. The human contribution will shift toward judgement, relationship management, ethical oversight, problem-solving, negotiation, and cross-functional coordination. In parallel, new or expanded roles are emerging in AI governance, data stewardship, workflow design, automation supervision, digital enablement, and change enablement.
For a quick, non-exhaustive, high-level view of exposure across common roles in HIQA/CORU-regulated healthcare and social care services, the risk matrix below points to a likely pattern of AI-enabled change: task automation first, role redesign second, and outright role loss mainly where work is highly routine, transactional, or heavily template-driven.
Grading method
- High exposure — a large share of tasks are routine, rules-based, document-heavy, schedulable, or easily standardised.
- Medium exposure — the role will not disappear, but a substantial part of the work will be redesigned, augmented, or moved into higher-value judgement and relationship tasks.
- Low exposure — work depends heavily on clinical judgement, human relationship, safeguarding, complex observation, or context-sensitive professional decision-making.
Risk matrix — administrative & coordination layer
| Role / function | Exposure | Why |
|---|---|---|
| Reception, switchboard, appointment booking | High | Highly repetitive, rules-based, and easily triaged by digital assistants and scheduling systems. |
| Data entry / records administration / scanning | High | Structured, repetitive information handling is highly automatable. |
| Payroll, invoicing, basic finance admin | High | Transactional processing and validation are prime automation targets. |
| Routine customer service / helpdesk / service desk | High | Scripted queries and first-line support are increasingly handled by AI systems. |
| Care roster coordination / basic scheduling | High | Predictable scheduling and coverage matching can be automated or agent-assisted. |
| Procurement admin / order processing | High | Rules-based approvals and matching tasks are strong automation candidates. |
| Quality admin / audit / compliance tracking | Medium | Reporting can be automated, but interpretation and escalation still need human oversight. |
| HR administration / recruitment support | Medium | Screening, shortlisting and correspondence can be automated; coordination judgement and candidate engagement remain human-led. |
| Social care support worker | Medium | Core care and relationship work remain human — documentation, handover, rostering and prompting tools will reshape the role. |
Risk matrix — clinical & professional roles
| Role / function | Exposure | Why |
|---|---|---|
| Healthcare assistant / care assistant | Medium | Direct care is resilient, but AI will affect documentation, observation support, and task planning. |
| Ward clerk / unit secretary | Medium | A mix of administrative and coordination tasks makes this role partly automatable and partly redesignable. |
| Medical secretary / dictation transcription | Medium | Transcription and document drafting are automatable, though clinical context and exception handling remain important. |
| Therapy assistant / rehabilitation assistant | Medium | Scheduling, measurement capture and documentation will change, but supervised human interaction remains central. |
| Staff nurse / registered nurse | Low–Medium | Direct care, risk assessment and relational judgement are resilient; AI will alter documentation, surveillance, handover and decision support. |
| Social worker | Low–Medium | Case recording and admin parts may be automated; statutory judgement, safeguarding and complex family work remain human-led. |
| Occupational therapist | Low | Holistic assessment, adaptation and functional reasoning are difficult to automate, though admin burdens will fall. |
| Physiotherapist | Low | Clinical reasoning and in-person intervention remain central; technology mainly augments assessment and programme tracking. |
| Speech and language therapist | Low | Complex interaction, assessment and therapeutic adjustment make the role relatively resilient. |
| Psychologist / psychotherapist | Low | Trust, nuance, formulation and human relationship are difficult to automate safely. |
| Medical practitioner / consultant | Low | Diagnostic support will expand, but final accountability, judgement and patient communication remain essential. |
| Nurse manager / service manager | Low–Medium | Workforce analytics and AI planning tools will change the role, but leadership, culture and accountability remain human. |
| Clinical governance / safeguarding / risk lead | Low | Oversight-heavy roles where AI is supportive, not substitutive, because of accountability and ethical risk. |
How to read the matrix
In HIQA- and CORU-regulated services, the highest exposure is usually not in front-line caring roles but in the administrative and coordination layer around care delivery — booking, records, payroll, compliance admin, scheduling, and routine communication. The literature also shows that many healthcare applications of AI fall into high-risk categories requiring strong oversight, which means front-line professional roles will be transformed rather than simply removed.
Likely role changes by pattern
- Administrative roles will shrink in volume but grow in digital complexity.
- Support roles will move toward supervision of systems, exception handling, and data quality.
- Professional roles will become more technology-enabled, with less time on documentation and more time on judgement, care, and relationship work.
- Management roles will increasingly rely on workforce analytics, automation planning, and change leadership.
Practical use in a regulated service
This matrix works best if applied task-by-task rather than only by job title. A role can be low exposure overall but still contain high-exposure tasks that should be redesigned now — especially documentation, scheduling, reporting, and routine correspondence. The most useful next step is to convert this into a workforce planning tool with columns for current headcount, automatable tasks, redeployment options, and training priority.
Immediate action areas
- Map roles by task content, not title alone.
- Identify high-exposure admin roles for redeployment and reskilling.
- Protect low-exposure clinical roles by reducing admin burden rather than cutting headcount.
- Update policies on AI use, record integrity, accountability, and human oversight.
- Build manager capability in change communication, consultation, and job redesign.
3. Competitive and talent implications
Organisations that move early can gain first-mover advantage through faster service, lower unit cost, better responsiveness, and stronger customer insight. The same advantage applies in the labour market: employers with clear AI strategies, credible development pathways, and modern work design will attract stronger candidates and retain scarce talent more effectively. The strategic risk is not only being technologically behind, but being seen as an unattractive employer by digitally fluent candidates.
Talent management success will depend on whether the organisation can reskill at pace, redesign jobs intelligently, and maintain trust during transition. Failure usually comes from treating AI as a pure efficiency project, cutting roles before redesigning work, or under-investing in the supervisors and line managers who must carry the change. Recruitment will increasingly favour candidates who combine technical literacy with adaptability, analytical thinking, resilience, curiosity, and learning agility.
4. People and OD implications
The workforce of the next five years will likely value speed, adaptability, ethical and psychologically safe environments, digital confidence, collaboration, and comfort with ambiguity. Staff retained will need to be good at working alongside machines, reviewing outputs critically, and learning continuously. Staff displaced from routine roles may need support into new pathways emphasising service, care, judgement, coordination, or technical support. For many organisations, the real OD challenge is not only capability, but identity: helping people understand what the organisation now values and what sort of entity it is evolving towards.
Management competence is already shifting from supervision-by-presence to leadership-by-context. Leaders need stronger skills in workforce analytics, change communication, ethical judgement, task redesign, and psychological safety, while traditional command-and-control styles become less effective in complex, fast-changing environments. The leadership style most likely to grow is adaptive, coaching-oriented, data-informed, and cross-functional, with visible emphasis on fairness, inclusion, and transparency.
5. Policies and readiness
Organisations should begin by reviewing workforce strategy, job architecture, skills frameworks, and change governance. Policies should cover AI use, data protection, accountability for automated decisions, quality assurance, human oversight, reskilling commitments, consultation, and ethical guardrails. Values also need to be explicit: if the organisation wants agility and innovation, it must also define what responsible use, dignity at work, and fair transition mean in practice.
Management competence should be developed in parallel with technology deployment — training leaders to assess work for automation potential, redesign tasks and roles, manage transition conversations, and use evidence rather than assumption when planning headcount or capability shifts. Organisations also need stronger scenario planning so they can model different futures rather than assuming one linear path.
6. Assessments and actions needed now
The most useful immediate actions are to map work rather than simply count roles. Start with a task-level review of where AI, automation, or interoperability can remove friction, then identify which jobs are fully exposed, partially exposed, or likely to grow in importance. A second priority is a skills-gap assessment focused on analytical thinking, digital fluency, AI literacy, change capability, and leadership readiness.
A practical launch plan might include
- A workforce segmentation review by role, task, and risk of automation.
- A critical roles analysis for retention, reskilling, and succession.
- An AI and data governance review.
- A manager capability audit.
- A reskilling and redeployment framework for vulnerable roles.
- A communication and consultation plan for staff and employee representatives.
- A 12–24 month scenario plan for market, service, and labour demand shifts.
