Role Definition
| Field | Value |
|---|---|
| Job Title | Junior Doctor — Foundation Year (F1/F2) |
| Seniority Level | Entry-Level (ONS SOC 2020: 2211) |
| Primary Function | Two-year rotational NHS training programme immediately after medical school. F1 (provisionally registered with GMC) and F2 (fully registered). Core daily work: ward rounds, clerking new patients, writing discharge summaries, prescribing medications, performing basic procedures under supervision (cannulation, catheterisation, ABGs, blood cultures), ordering and chasing investigations, handover, and patient communication. Nationally recruited via UKFPO, supervised by registrars and consultants, assessed against Foundation Programme Curriculum competencies. Approximately 17,000 in post across England at any time. |
| What This Role Is NOT | Not a consultant/attending physician (mid-to-senior, independently practising — scored separately at 63.6). Not a GP registrar or specialty trainee (ST3+). Not a physician associate or advanced clinical practitioner (different scope and licensing). Not a medical student (not yet registered with GMC). |
| Typical Experience | 5-6 years medical school (UK), 0-2 years post-qualification. UKMLA from 2025. No postgraduate exams required at this stage. |
Seniority note: This is the entry-level grade in UK medicine. Senior physicians (consultants) score higher (63.6) because they have greater clinical autonomy, more complex decision-making, and less time on automatable documentation. The seniority gap in medicine (11.6 points) is much smaller than in tech (Junior SW Dev 9.3 vs Senior SW Eng 55.4 = 46.1 points) because even junior doctors perform irreducible physical and interpersonal work.
Protective Principles + AI Growth Correlation
| Principle | Score (0-3) | Rationale |
|---|---|---|
| Embodied Physicality | 2 | Physical examination (auscultation, palpation, neurological assessment), bedside procedures (cannulation, catheterisation, ABGs, blood cultures), resuscitation participation. Clinical environments are semi-structured (wards, A&E) but unpredictable (acutely unwell patients, emergency calls). |
| Deep Interpersonal Connection | 2 | Explains diagnoses, breaks bad news (under supervision), communicates with distressed families, supports dying patients. Foundation doctors are often the most visible doctor on the ward — patients form real relationships with them. Trust matters, though it is not the sole value proposition. |
| Goal-Setting & Moral Judgment | 2 | Makes real-time clinical decisions under pressure — which patient to see first, when to escalate, when something feels wrong. Not yet setting treatment strategy (that is the consultant's role), but exercises genuine clinical judgment within their scope. Personal GMC accountability even at F1 level. Scored 2 not 3 because decision-making is supervised, not autonomous. |
| Protective Total | 6/9 | |
| AI Growth Correlation | 0 | AI adoption does not create or destroy demand for foundation doctors. Demand is driven by NHS training pipeline, population health needs, and workforce planning. AI may improve efficiency but cannot replace the training function — F1/F2 is an educational role, not just a service delivery role. |
Quick screen result: Protective 6/9 = Strong Green Zone signal. Proceed to confirm with task analysis.
Task Decomposition (Agentic AI Scoring)
| Task | Time % | Score (1-5) | Weighted | Aug/Disp | Rationale |
|---|---|---|---|---|---|
| Ward rounds (clerking, bedside assessment, physical examination) | 25% | 2 | 0.50 | AUGMENTATION | AI pre-populates patient summaries, flags overnight changes, suggests differentials. But the F1/F2 must physically examine the patient, present findings to the consultant, and document the plan. Licensed professional presence required at the bedside. |
| Clinical documentation (discharge summaries, progress notes, handover sheets) | 20% | 4 | 0.80 | DISPLACEMENT | Ambient AI scribes (Heidi Health, DAX in NHS pilots) generate notes from clinical conversations. Discharge summary drafting is a prime target — structured, template-driven, extractable from EPR data. F1s currently spend disproportionate time on documentation. AI output IS the deliverable; F1 reviews and signs. |
| Prescribing and medication management | 15% | 3 | 0.45 | AUGMENTATION | AI clinical decision support flags drug interactions, dose adjustments for renal/hepatic impairment, formulary compliance. Electronic prescribing systems increasingly suggest appropriate medications. But the prescribing decision and legal responsibility remain with the doctor. F1s must still assess the patient and exercise judgment on what to prescribe. Human-led, AI-accelerated. |
| Procedures under supervision (cannulation, catheterisation, ABGs, blood cultures) | 10% | 1 | 0.10 | NOT INVOLVED | Irreducible physical dexterity tasks in unstructured patient environments. Veins vary, patients move, anatomy is unpredictable. No robotic system performs bedside procedures in NHS ward settings. These are also essential training competencies — the educational purpose is as important as the service function. |
| Ordering and chasing investigations (bloods, imaging, results) | 10% | 4 | 0.40 | DISPLACEMENT | AI agents can order routine investigations based on clinical protocols, chase outstanding results, flag abnormals, and route them to the appropriate clinician. Much of this is currently clerical work done by F1s. Structured inputs, defined processes, verifiable outputs. Human reviews but does not need to be in the loop for every step. |
| Patient and family communication | 10% | 1 | 0.10 | NOT INVOLVED | Explaining a diagnosis to an anxious patient. Updating a worried family member. Discussing resuscitation decisions. Comforting a dying patient. The human connection IS the value. F1/F2 doctors are often the frontline for these conversations. |
| Referrals, handover, and coordination | 10% | 3 | 0.30 | AUGMENTATION | AI drafts referral letters, structures handover data (SBAR format), coordinates scheduling. But clinical judgment about what to refer, when to escalate, and how to prioritise remains human. Mixed: some sub-tasks agent-executable (letter drafting, data compilation), others require clinical context and interpersonal negotiation. |
| Total | 100% | 2.65 |
Task Resistance Score: 6.00 - 2.65 = 3.35/5.0
Displacement/Augmentation split: 30% displacement, 50% augmentation, 20% not involved.
Reinstatement check (Acemoglu): AI creates new tasks for foundation doctors: validating AI-generated discharge summaries, critically appraising AI diagnostic suggestions, learning to use and configure clinical decision support tools, auditing AI prescribing recommendations. The Foundation Programme Curriculum will evolve to include AI literacy competencies. Foundation doctors become the generation that learns medicine with AI from day one — a transformative shift in medical education, not a displacement signal.
Evidence Score
| Dimension | Score (-2 to 2) | Evidence |
|---|---|---|
| Job Posting Trends | 2 | UKFPO applicants grew from 8,137 (2019) to 11,205 (2025) — 38% increase in 6 years. Foundation Programme is oversubscribed, not contracting. NHS Long Term Workforce Plan (2023) targets 15,000 additional medical training places by 2031. Zero displacement signal — the pipeline is expanding. |
| Company Actions | 1 | NHS trusts deploying AI tools (Heidi Health ambient scribes, DAX pilots, Epic AI modules) to support junior doctors, not replace them. NHS 10-Year Health Plan (2025) frames AI as "freeing up clinician time." No trust cutting junior doctor posts citing AI. But deployment is augmentation-focused, not headcount-expanding. |
| Wage Trends | 1 | Basic F1 salary GBP 38,831 (2025-26). Average resident doctor pay approximately GBP 54,300 after 28.9% increase over 3 years (GOV.UK). Real-terms recovery after years of pay erosion (BMA estimated 26% real-terms cut 2008-2023). Now growing above inflation following industrial action settlement. |
| AI Tool Maturity | 0 | Ambient AI scribes (Heidi, DAX) in NHS trust pilots — cutting documentation time 51% in trials. Clinical decision support in EPRs. NHS AI roadmap targets imaging and coding AI in year one, agentic AI by 2026-27. But no tool can independently examine patients, prescribe, or perform procedures. Tools in pilot/early adoption at most trusts — unclear headcount impact. |
| Expert Consensus | 1 | NHS England, Royal College of Physicians, GMC, Health Education England consensus: AI augments junior doctors. Topol Review (2019): AI will transform clinical roles, not replace them. Parliamentary answer (Dec 2025): AI supporting "routine administrative tasks and clinical decision-making." No credible source predicts displacement of training-grade doctors. |
| Total | 5 |
Barrier Assessment
Reframed question: What prevents AI execution even when programmatically possible?
| Barrier | Score (0-2) | Rationale |
|---|---|---|
| Regulatory/Licensing | 2 | GMC provisional registration (F1) or full registration (F2) required. Medical degree from GMC-approved school. UK Medical Licensing Assessment (UKMLA) mandated from 2025. Annual Review of Competence Progression (ARCP). No regulatory pathway exists for AI to hold GMC registration or practise medicine independently in the UK. |
| Physical Presence | 2 | Physical examination, bedside procedures, resuscitation, ward-based patient assessment all require hands-on presence. F1/F2 work is predominantly ward-based in unstructured clinical environments — patients deteriorate unpredictably, require physical assessment, and need procedural interventions that no robotic system can deliver in NHS settings. |
| Union/Collective Bargaining | 1 | BMA Junior Doctor Committee represents foundation doctors with active collective bargaining. Historic industrial action 2023-2024 (longest junior doctor strikes in NHS history). Pay deal reached 2024. BMA actively protects terms, conditions, and training quality. Stronger union protection than most entry-level roles. |
| Liability/Accountability | 2 | Even provisionally registered F1 doctors bear personal professional accountability under GMC fitness to practise framework. Gross negligence manslaughter prosecutions (Dr Bawa-Garba case, 2018) demonstrate criminal liability reaches foundation doctors. Medical Defence Organisation (MDU/MPS) membership required. No AI system can bear GMC registration or criminal liability. |
| Cultural/Ethical | 1 | Patients expect to see a human doctor. The doctor-patient relationship — including at foundation level — carries cultural weight. But foundation doctors are known to be trainees under supervision; cultural trust is weaker than for consultants. Patients may accept AI assistance more readily when they know the doctor is junior. Scored 1 not 2 for this reason. |
| Total | 8/10 |
AI Growth Correlation Check
Confirmed 0 (Neutral). AI adoption does not inherently create or destroy demand for foundation doctors. The Foundation Programme exists to train the next generation of UK doctors — its size is determined by medical school output, workforce planning, and population health needs, not by AI adoption rates. AI may improve foundation doctor efficiency (fewer hours on documentation, faster investigation ordering), but this time is reinvested in clinical learning and patient care, not in reducing headcount. Not Accelerated Green — no recursive AI dependency.
JobZone Composite Score (AIJRI)
| Input | Value |
|---|---|
| Task Resistance Score | 3.35/5.0 |
| Evidence Modifier | 1.0 + (5 x 0.04) = 1.20 |
| Barrier Modifier | 1.0 + (8 x 0.02) = 1.16 |
| Growth Modifier | 1.0 + (0 x 0.05) = 1.00 |
Raw: 3.35 x 1.20 x 1.16 x 1.00 = 4.6632
JobZone Score: (4.6632 - 0.54) / 7.93 x 100 = 52.0/100
Zone: GREEN (Green >=48, Yellow 25-47, Red <25)
Sub-Label Determination
| Metric | Value |
|---|---|
| % of task time scoring 3+ | 55% |
| AI Growth Correlation | 0 |
| Sub-label | Green (Transforming) — >=20% task time scores 3+ |
Assessor override: None — formula score accepted. The 52.0 score places the role 4.0 points above the Green/Yellow boundary, which is appropriate. The task resistance (3.35) is lower than the mid-to-senior physician (3.60) because foundation doctors spend proportionally more time on automatable documentation and investigation ordering. The 8/10 barrier score provides meaningful structural protection (16% boost) that reflects the reality of GMC licensing, physical presence, and personal liability. No override needed.
Assessor Commentary
Score vs Reality Check
The 52.0 AIJRI places this role 4.0 points above the Green/Yellow boundary — not deeply Green but honestly positioned. The task resistance (3.35) is notably lower than the mid-to-senior physician (3.60) because F1/F2 doctors spend disproportionate time on documentation and clerical tasks that AI is already automating. The 8/10 barrier score does heavy lifting — without barriers, the AIJRI would drop to approximately 44.8 (Yellow). This barrier dependency is real and well-founded: GMC registration, physical clinical presence, and personal criminal liability are structural protections that will not erode on any foreseeable timeline. The label is honest but should be understood as barrier-supported Green.
What the Numbers Don't Capture
- Training function as protection. The Foundation Programme is an educational pipeline, not just a service delivery model. Even if AI could do 100% of F1 clinical tasks, the training role would persist because the NHS needs to produce consultants, GPs, and specialists. You cannot become a senior doctor without being a junior doctor first. This is an irreducible structural protection that task analysis cannot score.
- Bimodal distribution. F1/F2 work splits sharply between automatable administrative tasks (documentation, ordering, chasing — 30% at score 4) and irreducible clinical tasks (examination, procedures, patient communication — 20% at score 1). The average masks the split. The administrative half of the role is transforming fast; the clinical half is untouched.
- Pay erosion and retention crisis. UK junior doctor pay fell approximately 26% in real terms between 2008 and 2023 (BMA data). The 2023-2024 industrial action and subsequent settlement partially corrected this, but retention remains fragile. Approximately 40% of F2 doctors do not enter UK specialty training immediately — many take career breaks, locum, or emigrate. This is a workforce sustainability issue, not an AI issue, but it affects the evidence score.
- NHS digital maturity variation. AI tool adoption varies enormously across NHS trusts. Some (Royal Devon, Alder Hey) are piloting ambient AI scribes; others still use paper notes and fax machines. The national average masks trusts where F1/F2 work is entirely pre-digital. AI transformation will be uneven and slow in many settings.
Who Should Worry (and Who Shouldn't)
No foundation doctor should worry about AI displacing their role. The GMC registration, physical clinical work, and training pipeline make this one of the most structurally protected entry-level roles in any profession. The "Transforming" label means the daily workflow is changing — less time writing discharge summaries, more time with patients — which is unambiguously positive for junior doctors. Most protected: F1/F2s in acute medicine, surgery, A&E, and paediatrics where physical examination and procedures dominate the day. Slightly more exposed to workflow change (not displacement): F1/F2s in outpatient-heavy rotations or specialties with high documentation loads (general medicine, care of the elderly) where AI scribes will transform daily work fastest. The single biggest factor: whether your trust has invested in digital infrastructure. Foundation doctors in digitally mature trusts will see AI tools first — and will gain the AI literacy that becomes essential for the rest of their career.
What This Means
The role in 2028: Foundation doctors will use AI ambient documentation as standard in digitally mature trusts — discharge summaries drafted automatically, progress notes generated from ward round conversations, handover sheets compiled by AI agents. Investigation ordering will be partially automated with AI suggesting appropriate tests based on clinical context. Prescribing decision support will be more sophisticated, flagging not just interactions but suggesting evidence-based treatment pathways. But the foundation doctor will still examine every patient, perform every cannula, attend every cardiac arrest call, and explain every diagnosis to every worried family. The administrative burden that currently consumes 30-40% of F1/F2 time will shrink — and that time will be reinvested in clinical learning and patient care.
Survival strategy:
- Embrace AI documentation tools as they arrive at your trust — learn to validate and edit AI-generated notes efficiently rather than writing everything from scratch
- Invest heavily in the irreducible clinical skills: physical examination, procedural competence, clinical reasoning under uncertainty, and patient communication — these are the skills AI cannot replicate and that define your long-term career value
- Build AI literacy during foundation training — understand how clinical decision support works, its limitations, and how to critically appraise AI-generated recommendations, as this will differentiate you throughout specialty training and beyond
Timeline: 15-25+ years, if ever. Constrained by GMC registration requirements (5-6 years medical school + 2 years foundation as minimum pathway), personal criminal liability for clinical decisions (no AI can hold GMC registration), regulatory mandates (no UK regulatory pathway for autonomous AI clinical practice), and the irreducible training function (the NHS must produce future consultants and GPs through supervised junior doctor rotations).
