Nurse Midwives

29-1161.00
Median wage $134,040/yr7,920 employed (US)Rank #715 of 923 scored · top 77% by substitution

Diagnose and coordinate all aspects of the birthing process, either independently or as part of a healthcare team. May provide well-woman gynecological care. Must have specialized, graduate nursing education.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution18
Exposure19
Augmentation62

Substitution — the headline: capability discounted by cost, barriers and adoption.

Exposure — technical capability alone, regardless of whether anyone deploys it.

Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.

Tasks on the substitution scale

21 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

0%

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%17

panel mean rating 1.7/5 → substitution pressure 17/100

Technical feasibility todayw 20%21

panel mean rating 1.9/5 → substitution pressure 21/100

Cost vs. human wagew 15%20

panel mean rating 1.8/5 → substitution pressure 20/100

Adoption barriersw 20%inverted — strong barriers lower the score15

panel mean rating 4.4/5 (barrier strength) → substitution pressure 15/100

Sector adoption velocityw 10%22

panel mean rating 1.9/5 → substitution pressure 22/100

Task breakdown (21 tasks)

Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.

Read current literature, talk with colleagues, or participate in professional organizations or conferences to keep abreast of developments in midwifery.

64

CI 56–72 · exposure 62 · augmentation 88 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Healthcare organizations show moderate uptake of AI literature tools and knowledge management systems, with pilots common in academic medical centers, but deployment in routine clinical practice remains uneven and adoption is slower than in information-heavy sectors like finance or law.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially nursing/midwifery, has been slower than tech or finance sectors to adopt AI tools for continuing education and professional development workflows.
Augmentation potentialclaude-haiku-4-5-202510015/5AI dramatically augments a midwife's ability to stay current by rapidly filtering, summarizing, and curating dozens of new publications and conference insights, allowing the human expert to focus on critical appraisal and application to clinical practice rather than raw information triage.
Augmentation potentialclaude-sonnet-54/5AI tools can efficiently curate, summarize, and highlight relevant new research and guidelines, meaningfully saving time and improving currency of knowledge while the midwife remains the one engaging with peers and organizations.
Task automatabilityclaude-haiku-4-5-202510014/5AI can comprehensively scan and summarize current midwifery literature, synthesize peer-reviewed research, and extract key developments from conferences and professional organization materials with minimal manual curation, easily achieving 50% time savings on the information-gathering phase while maintaining quality.
Task automatabilityclaude-sonnet-53/5AI can summarize literature and surface relevant developments, but the task also includes networking, attending conferences, and professional participation that require human presence and social engagement.-The reading/synthesis portion is highly automatable, but the full task bundle is not.
Adoption barriersclaude-haiku-4-5-202510012/5No licensing or regulatory requirement mandates human reading of literature; professional society membership and conference access may carry some organizational friction, but these are weak adoption barriers relative to the task's discretionary nature.
Adoption barriersclaude-sonnet-52/5No licensing requirement mandates a human read literature personally, though continuing education/CME requirements sometimes mandate documented human participation in professional development activities.
Cost vs. human wageclaude-haiku-4-5-202510014/5LLM inference and integration costs are negligible compared to the loaded wage of a nurse midwife spending hours reading journals and attending conferences, making AI-assisted literature review an order of magnitude cheaper per information-unit extracted.
Cost vs. human wageclaude-sonnet-54/5AI-based literature scanning and summarization is very cheap compared to a midwife's time spent reading journals, though the networking/CME components still require human cost.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed systems like GPT-4, specialized medical AI tools, and journal aggregation platforms reliably perform literature scanning and summarization in production; however, human judgment remains valuable for filtering relevance and depth, so this falls short of fully autonomous end-to-end execution.
Technical feasibility todayclaude-sonnet-53/5AI literature summarization and alert tools (e.g., research digest apps, LLM-based summarizers) are deployed and used in clinical fields, but conference participation and colleague discussion aspects have no AI substitute in production.

Document findings of physical examinations.

58

CI 45–71 · exposure 62 · augmentation 88 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Healthcare is digitizing rapidly, and EHR voice-to-note adoption is growing in hospitals and maternity clinics, but midwifery is smaller and more fragmented than general nursing, with slower tech adoption in community-based and rural settings. Pilots are common but production integration remains middling.
Sector adoption velocityclaude-sonnet-53/5Healthcare documentation AI is spreading via ambient scribe products, but overall healthcare adoption remains slower than in finance or tech due to compliance and integration hurdles.
Augmentation potentialclaude-haiku-4-5-202510015/5AI documentation assistance substantially augments midwife productivity by eliminating manual note-writing, reducing cognitive load, and freeing time for direct patient care. Clinicians retain full oversight, judgment, and signature authority while benefiting from rapid, accurate documentation drafts.
Augmentation potentialclaude-sonnet-54/5AI scribes and transcription tools meaningfully speed up documentation of exam findings while the midwife retains responsibility for accuracy and clinical judgment.
Task automatabilityclaude-haiku-4-5-202510014/5Current AI systems can reliably document and transcribe physical examination findings from voice or structured templates with high accuracy. Speech-to-text and clinical documentation templates enable near-complete automation with minimal human editing, achieving >50% time savings. However, nuanced clinical judgment about what to emphasize or follow up on may still require human review.
Task automatabilityclaude-sonnet-53/5AI can draft clinical documentation from dictation or structured inputs, saving significant time, but requires clinician review and correction to ensure accuracy for the medical record.
Adoption barriersclaude-haiku-4-5-202510014/5Significant regulatory and liability barriers exist: clinicians must review and sign off on documentation, and malpractice risk means human accountability cannot be fully delegated. Legal requirements in most jurisdictions mandate that a licensed clinician verify and attest to the medical record, preventing full automation.
Adoption barriersclaude-sonnet-53/5Medical records require accuracy and legal accountability from a licensed provider, so a human midwife must review and sign off, though AI-assisted drafting itself faces no direct prohibition.
Cost vs. human wageclaude-haiku-4-5-202510015/5Automated speech-to-text and note generation costs a small fraction of the loaded midwife wage (~$150/hour + overhead), with API costs typically <$1 per note after integration. AI is at least 50–100× cheaper per documented exam.
Cost vs. human wageclaude-sonnet-53/5AI scribe subscriptions are cheaper than dedicated transcription staff but still require licensing fees and clinician review time, keeping costs roughly comparable rather than order-of-magnitude cheaper.
Technical feasibility todayclaude-haiku-4-5-202510014/5Mature clinical documentation AI products (EHR voice-to-note systems, medical transcription platforms) are deployed in production across healthcare systems and reliably generate documentation from examination findings. Accuracy on standard findings is high, though integration with specific clinical workflows varies.
Technical feasibility todayclaude-sonnet-53/5Ambient clinical documentation tools (e.g., DAX, Nuance) are deployed in some obstetric/midwifery practices, but adoption in midwifery specifically is narrower than in general primary care.

Document patients' health histories, symptoms, physical conditions, or other diagnostic information.

41

CI 32–50 · exposure 42 · augmentation 75 · importance 5.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Healthcare has begun adopting AI scribe and documentation tools, with increasing pilots in hospitals and clinics, but rollout remains uneven and adoption is slower than in tech/finance sectors due to regulatory caution, integration complexity, and clinician skepticism about accuracy and liability.
Sector adoption velocityclaude-sonnet-53/5Healthcare has been a middling adopter of AI documentation tools; ambient scribes are gaining traction but broad, deep adoption across midwifery practices specifically lags larger health systems.
Augmentation potentialclaude-haiku-4-5-202510014/5AI transcription, summarization, and template-based documentation tools meaningfully assist midwives by reducing manual typing and organizing data, allowing practitioners to focus on patient interaction and clinical judgment rather than administrative burdens.
Augmentation potentialclaude-sonnet-54/5AI-powered transcription and note-generation tools meaningfully speed up documentation and reduce clinician burden while the midwife remains responsible for verifying and finalizing clinical content.
Task automatabilityclaude-haiku-4-5-202510012/5Documenting health histories requires capturing nuanced patient narratives, interpreting clinical observations, and integrating information into structured medical records. While AI can assist with transcription and template population, it cannot reliably extract and synthesize the complex, context-dependent clinical judgments required for complete and accurate documentation without substantial human review.
Task automatabilityclaude-sonnet-53/5AI scribes and ambient documentation tools can transcribe and structure clinical encounters into notes, but capturing accurate symptoms and physical exam findings still requires human verification and correction for clinical accuracy.
Adoption barriersclaude-haiku-4-5-202510014/5Medical documentation carries high liability risk, and legal and regulatory frameworks (HIPAA, state scope-of-practice laws, malpractice standards) generally require a licensed practitioner to authenticate and sign clinical records. Many healthcare systems maintain strict policies that the responsible clinician must review and certify all documentation.
Adoption barriersclaude-sonnet-53/5Medical documentation must be reviewed and attested by a licensed provider, and HIPAA/liability concerns create moderate barriers to full automation, though AI-assisted drafting is increasingly accepted.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI scribe services cost roughly $15–40 per encounter or $3–5k/month subscriptions; midwives and nurses earn $50–80/hour loaded cost. The AI reduces typing time significantly but integration, error correction, and regulatory compliance still consume substantial human oversight, yielding only modest overall savings.
Cost vs. human wageclaude-sonnet-53/5Ambient scribe subscriptions are cheaper than dedicated human scribes but still require licensing fees plus clinician review time, making savings moderate rather than order-of-magnitude.
Technical feasibility todayclaude-haiku-4-5-202510013/5Scribe AI and voice-to-text documentation systems exist in production (e.g., Ambient by Nuance, Aura), but they still require material physician/provider review, correction, and sign-off. These tools handle structured data entry well but struggle with nuanced clinical interpretation and legal accountability.
Technical feasibility todayclaude-sonnet-53/5Ambient AI documentation products (e.g., DAX Copilot, Nuance, Abridge) are deployed in some clinical settings including midwifery/OB practices, but adoption is uneven and error correction by clinicians remains standard practice.

Write information in medical records or provide narrative summaries to communicate patient information to other health care providers.

39

CI 32–45 · exposure 42 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Healthcare organizations are piloting AI-assisted documentation widely, but adoption of AI-generated summaries without clinician review remains limited due to liability and accuracy concerns. Most deployments function as augmentation tools requiring human validation rather than replacement.
Sector adoption velocityclaude-sonnet-53/5Healthcare overall is a middling-adoption sector with growing pilots of AI scribes and documentation tools but uneven and cautious rollout across specialties like midwifery.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist by extracting relevant data, suggesting structured summaries, and reducing transcription burden, allowing clinicians to focus on clinical content and decision-making. This augmentation is actively used in many EHR implementations where clinicians retain full authorship and accountability.
Augmentation potentialclaude-sonnet-54/5AI scribing and summarization tools meaningfully speed up drafting of notes and handoff summaries, letting midwives focus more on patient interaction while still reviewing and finalizing entries.
Task automatabilityclaude-haiku-4-5-202510012/5AI can draft summaries and populate templates from clinical data, but the task requires clinical judgment about what information is clinically significant, accurate interpretation of patient context, and legal accountability for medical record accuracy. Current systems cannot reliably perform this end-to-end with sufficient quality for clinical safety.
Task automatabilityclaude-sonnet-53/5AI can draft clinical notes and summaries from structured data or dictation with substantial time savings, but a licensed clinician must verify accuracy and clinical judgment content before finalizing.time-saving is real but not full end-to-end automation of the underlying documentation responsibility.
Adoption barriersclaude-haiku-4-5-202510014/5Medical records must be legally accurate and signed by the responsible clinician; liability for documentation errors falls on the nurse midwife. Regulatory (HIPAA, state nursing boards) and malpractice concerns create strong organizational and legal barriers to full automation without human oversight and sign-off.
Adoption barriersclaude-sonnet-54/5Medical record entries carry legal and clinical liability, requiring the credentialed provider to review, attest to, and sign the documentation, creating a substantial oversight and accountability barrier.
Cost vs. human wageclaude-haiku-4-5-202510012/5Implementation requires integration with EHR systems, oversight by licensed clinicians, and significant validation effort. The all-in cost of AI-assisted documentation remains comparable to or exceeds direct human charting, especially when accounting for necessary review cycles.
Cost vs. human wageclaude-sonnet-53/5AI scribe tools reduce documentation time meaningfully, but licensing fees, EHR integration, and required clinician review keep costs roughly comparable rather than order-of-magnitude cheaper.
Technical feasibility todayclaude-haiku-4-5-202510013/5Clinical documentation tools and AI-assisted drafting exist in production EHR systems, but they typically generate templates or suggestions that require substantial human review and correction. Reliability remains limited for complex cases, and clinician verification is standard, not exception.
Technical feasibility todayclaude-sonnet-53/5Ambient documentation and clinical note-generation products (e.g., DAX, Nuance, Abridge-style tools) are deployed in some health systems, but adoption in midwifery/obstetric settings specifically is narrower and error correction remains routine.

Plan, provide, or evaluate educational programs for nursing staff, health care teams, or the community.

28

CI 25–30 · exposure 25 · augmentation 63 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare organizations adopt AI for administrative tasks but remain cautious about core educational program design. Most adoption is exploratory or limited to content supplementation; production-scale replacement of program planning and evaluation remains rare in nursing education.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially nursing and midwifery, has been slower to adopt AI-driven training tools compared to sectors like tech or finance, though pilots in e-learning content generation exist.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist with generating content outlines, identifying evidence-based practices, scheduling logistics, and compiling evaluation data. However, the human educator remains essential for curriculum strategy, learning outcome design, and program stakeholder alignment, making this a moderate augmentation scenario.
Augmentation potentialclaude-sonnet-54/5AI tools can meaningfully assist in drafting curricula, generating quizzes, summarizing best practices, and creating materials, significantly speeding up program preparation even though delivery and evaluation remain human-led.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with content drafting, scheduling, and logistical coordination, but curriculum design, assessment strategy, and program evaluation require domain expertise, stakeholder input, and judgment that AI cannot reliably provide end-to-end. The task involves complex organizational and pedagogical decisions that fall far short of the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5AI can help draft curricula or training materials, but planning, delivering, and evaluating educational programs for staff or community requires contextual judgment, live facilitation, and interpersonal engagement that current AI cannot fully replace.
Adoption barriersclaude-haiku-4-5-202510014/5Healthcare education programs typically require accreditation standards, regulatory compliance, and institutional governance review. Nursing staff and learners often expect human instructors and mentors; clinical judgment in program design and evaluation demands credentialed professionals, creating organizational and legal friction against full automation.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement for the education task itself, but organizational expectations for clinical expertise and accountability for accuracy in health education create moderate friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-assisted content generation is relatively cheap, but the required human oversight—clinical review, pedagogical design, stakeholder coordination, and outcomes evaluation—makes the all-in cost comparable to or higher than hiring qualified nursing educators or instructional designers.
Cost vs. human wageclaude-sonnet-52/5AI can cheaply generate drafts of educational content, but the human effort in program design, live teaching, and evaluation dominates the cost, keeping overall savings modest.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI tools exist for generating educational materials and outlines, no deployed product reliably performs complete program planning and evaluation in healthcare education contexts. Existing systems lack the clinical credibility, compliance knowledge, and stakeholder alignment needed for production use in regulated healthcare settings.
Technical feasibility todayclaude-sonnet-52/5Products like AI-assisted content generators and LMS platforms exist for drafting materials, but no deployed product reliably plans and evaluates full educational programs in clinical/community settings autonomously.

Educate patients and family members regarding prenatal, intrapartum, postpartum, newborn, or interconception care.

27

CI 25–29 · exposure 25 · augmentation 75 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare remains a laggard in AI automation of clinical tasks, and midwifery is small-scale and regulation-intensive. Although some obstetric practices use AI for appointment reminders and generic education materials, real displacement of the education task itself remains minimal; pilots outnumber production deployments.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially maternal care, is a historically slow-adopting sector for patient-facing AI due to regulatory caution, liability concerns, and the high-touch nature of care.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist midwives by generating draft educational materials, summarizing clinical guidelines, creating personalized handouts, and fielding routine questions before or after visits. These tools raise productivity while the midwife retains responsibility for interpretation, cultural fit, and clinical judgment, making augmentation a strong use case.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist midwives by generating patient handouts, answering routine FAQs, and providing multilingual educational materials, freeing clinician time for higher-value interaction.
Task automatabilityclaude-haiku-4-5-202510012/5Patient education requires significant interpersonal judgment, tailoring to individual needs, and real-time responsiveness to questions and concerns. While AI could generate educational materials or scripts, the nuanced, adaptive counseling and relationship-building core to this task cannot be reliably automated end-to-end at equal quality without substantial human oversight.
Task automatabilityclaude-sonnet-52/5AI can generate patient education content and answer general questions, but delivering personalized education requires physical assessment, relationship-building, and real-time responsiveness that current systems cannot fully replace end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5The patient-contact requirement is central to prenatal and postpartum education, and malpractice and licensure frameworks vest accountability in licensed midwives. Regulatory bodies expect qualified professionals to provide personalized counseling; substituting AI alone creates liability and violates standard-of-care expectations in obstetric settings.
Adoption barriersclaude-sonnet-54/5Midwifery education involves clinical judgment and often requires licensed practitioner involvement for safety-critical guidance, with liability concerns around incorrect prenatal/postpartum advice creating strong barriers to full substitution.
Cost vs. human wageclaude-haiku-4-5-202510012/5Educational AI systems (chatbots, video platforms) have low inference costs but require significant content curation, clinical validation, and integration overhead. When accounting for oversight, personalization, and the need to supplement rather than replace human educators, the all-in cost approaches or exceeds the nurse midwife's loaded wage for this task.
Cost vs. human wageclaude-sonnet-53/5AI-generated educational materials are cheap to produce, but integration into clinical workflow, oversight for medical accuracy, and liability review keep costs comparable rather than dramatically lower.
Technical feasibility todayclaude-haiku-4-5-202510012/5Chatbots and educational platforms exist for prenatal information delivery, but no deployed product reliably performs full patient education (prenatal, intrapartum, postpartum, newborn, and interconception) with the clinical judgment, cultural sensitivity, and real-time adaptability required in clinical practice. Products are narrow in scope and lack production validation in midwifery settings.
Technical feasibility todayclaude-sonnet-52/5Chatbots and patient portals exist for general health education, but no deployed product reliably delivers comprehensive, personalized prenatal/postpartum counseling in clinical practice at scale.

Conduct clinical research on topics such as maternal or infant health care, contraceptive methods, breastfeeding, and gynecological care.

25

CI 25–25 · exposure 25 · augmentation 75 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Clinical research adoption of AI is concentrated in data analysis and secondary tasks; primary research design and execution remain human-driven. The nursing/midwifery field lags compared to software or finance, and research itself is not a high-volume task in most clinical settings.
Sector adoption velocityclaude-sonnet-52/5Healthcare research settings adopt AI tools for literature synthesis and data analysis at a moderate pace, but clinical research protocols and hands-on midwifery research remain slow to digitize due to regulatory and ethical constraints.
Augmentation potentialclaude-haiku-4-5-202510014/5AI substantially augments researcher productivity through literature synthesis, systematic review support, statistical analysis, manuscript drafting, and protocol optimization. These tools meaningfully accelerate research workflows while the midwife researcher retains full control and judgment over study direction and integrity.
Augmentation potentialclaude-sonnet-54/5AI substantially aids literature reviews, statistical analysis, hypothesis generation, and manuscript drafting, meaningfully boosting productivity for nurse midwives conducting clinical research while they retain oversight and decision-making.
Task automatabilityclaude-haiku-4-5-202510012/5Clinical research involves design, recruitment, data collection, analysis, and interpretation that require domain expertise and human judgment. AI can assist with literature reviews, data analysis, and manuscript drafting, but cannot autonomously design rigorous studies, obtain informed consent, or conduct the clinical observations that define the research process.
Task automatabilityclaude-sonnet-52/5AI can assist with literature review, data analysis, and drafting protocols, but designing, conducting, and interpreting original clinical research requires human expertise, ethics oversight, and physical/patient interaction that current AI cannot fully replace.
Adoption barriersclaude-haiku-4-5-202510014/5Clinical research is heavily regulated (IRB approval, informed consent, data protection laws, publishing standards). Liability for research integrity, ethical compliance, and adverse outcomes falls on licensed clinicians who must oversee and sign off on studies, creating strong legal and regulatory barriers to full automation.
Adoption barriersclaude-sonnet-54/5Clinical research involving human subjects requires IRB approval, licensed clinician oversight, and regulatory compliance (e.g., HHS, FDA), creating strong institutional and legal barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5While AI can reduce costs in specific phases (literature review, data processing), the full research cycle requires highly paid clinical researchers and midwives. The all-in cost of AI infrastructure, validation, and human oversight remains comparable to or exceeds the cost of skilled human researchers.
Cost vs. human wageclaude-sonnet-52/5AI can cheaply assist with search, summarization, and statistics, but the overall research process still requires costly human labor for study design, IRB compliance, patient recruitment, and clinical judgment, keeping the aggregate cost comparable to human-led research.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product autonomously conducts clinical research end-to-end. AI tools exist for individual components (statistical analysis, literature mining), but the coordination, ethical oversight, and clinical decision-making required remain human-dependent; research-stage capabilities only.
Technical feasibility todayclaude-sonnet-52/5AI literature-review and data-analysis tools are used in research support roles today, but no deployed product independently conducts clinical research studies at scale in production.

Develop and implement individualized plans for health care management.

23

CI 20–25 · exposure 25 · augmentation 75 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of AI for clinical decision-making remains cautious and fragmented; while some hospitals pilot decision support, production deployment of autonomous care planning is rare, and midwifery-specific AI tools have minimal market penetration due to clinical, regulatory, and liability concerns.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially direct clinical care in midwifery, adopts AI slowly due to regulatory, liability, and safety-critical constraints compared to information/finance sectors.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully augment midwives by rapidly synthesizing evidence, patient history, and guideline recommendations into draft plans, allowing clinicians to focus on personalization, communication, and clinical judgment while significantly improving plan completeness and evidence alignment.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist by synthesizing patient data, flagging risk factors, and drafting plan templates, letting the midwife focus judgment and patient interaction.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in drafting care plans by synthesizing clinical guidelines and patient data, the task requires integrating complex medical judgment, patient values, and real-time clinical reasoning that current systems cannot reliably perform end-to-end without substantial human oversight and modification.
Task automatabilityclaude-sonnet-52/5Care plan drafting can be partially templated, but individualized clinical judgment integrating patient history, risk factors, and preferences for pregnancy/birth management resists full automation to the equal-quality bar.
Adoption barriersclaude-haiku-4-5-202510014/5Nursing and midwifery are heavily regulated professions; care plan development and implementation are legal and professional responsibilities of licensed practitioners, and liability for adverse outcomes falls on the human clinician, creating strong regulatory and legal barriers to full automation.
Adoption barriersclaude-sonnet-55/5Care plan development and implementation for pregnancy/health management legally requires a licensed midwife or clinician's professional judgment and sign-off, with high liability exposure.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-powered decision support tools require significant infrastructure, clinical validation, and human oversight costs that approach or meet the cost of direct midwife labor, especially when accounting for liability and integration into clinical workflows.
Cost vs. human wageclaude-sonnet-52/5AI drafting assistance is cheap, but the human clinical judgment, exam, and accountability required keep overall cost comparable to human-led planning rather than dramatically cheaper.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some clinical decision support systems exist to generate plan suggestions, but they operate in narrow domains and require extensive human validation; no deployed products can independently develop and implement comprehensive individualized health care management plans for the diversity of patient presentations midwives encounter.
Technical feasibility todayclaude-sonnet-52/5Clinical decision-support tools exist that suggest care pathways, but no deployed product autonomously creates and implements individualized midwifery care plans in production without clinician authorship.

Order and interpret diagnostic or laboratory tests.

23

CI 20–25 · exposure 25 · augmentation 75 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare sectors show cautious, slow adoption of autonomous clinical decision-making due to liability and regulatory concerns. While EHR-integrated tools exist, deep production deployment of AI ordering systems remains limited, particularly in midwifery where human judgment and accountability are paramount.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially maternal care, adopts AI decision-support slowly due to regulatory, liability, and safety concerns, with pilots more common than widespread production use.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist midwives by flagging abnormal results, suggesting differential diagnoses, and providing evidence-based interpretation guides. This augmentation raises diagnostic accuracy and speed while the midwife retains ordering and clinical judgment authority, creating substantial productivity gains.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist by flagging abnormal results, summarizing trends, and suggesting differential considerations, improving efficiency while the midwife retains final interpretive authority.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in interpreting some test results (e.g., flagging abnormal values), the task requires clinical judgment about which tests to order based on individual patient context, pregnancy stage, and risk factors. Current AI systems cannot reliably make autonomous ordering decisions at the complexity level required for prenatal/obstetric care.
Task automatabilityclaude-sonnet-52/5Interpreting labs in the context of clinical judgment and patient-specific factors requires medical reasoning and accountability that current AI cannot fully replicate end-to-end without clinician oversight.
Adoption barriersclaude-haiku-4-5-202510014/5Ordering diagnostic tests is typically a licensed scope-of-practice function that may require physician oversight in some jurisdictions; liability for incorrect ordering or interpretation falls on the healthcare provider. Regulatory frameworks and malpractice concerns create strong barriers to autonomous AI ordering without human sign-off.
Adoption barriersclaude-sonnet-55/5Ordering and interpreting diagnostic tests is a licensed clinical act requiring a credentialed provider's authority and legal accountability, making this a hard-barrier task.
Cost vs. human wageclaude-haiku-4-5-202510012/5Integration costs for clinical decision support in midwifery are substantial, and the task involves relatively quick human decision-making. AI cost savings would be modest compared to the time a midwife spends on this task, making the cost ratio unfavorable for full automation.
Cost vs. human wageclaude-sonnet-52/5AI-assisted lab interpretation tools add cost on top of required clinician review, so total cost is not substantially cheaper than the clinician's time, especially given liability needs.
Technical feasibility todayclaude-haiku-4-5-202510012/5Clinical decision support tools exist to aid interpretation, but no deployed product reliably performs autonomous test ordering and interpretation for midwifery practice. Interpretation aids are available but orders require human clinical judgment, and liability concerns limit end-to-end automation in production settings.
Technical feasibility todayclaude-sonnet-52/5Some clinical decision-support tools flag abnormal values or suggest interpretations, but no deployed product independently orders and interprets diagnostic tests reliably in midwifery practice.

Explain procedures to patients, family members, staff members or others.

17

CI 9–25 · exposure 17 · augmentation 63 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Maternity care remains a high-touch, relationship-centered field with strong regulatory oversight and liability concerns. Healthcare adoption of AI for patient-facing explanation is lagging, and obstetric/midwifery settings show particularly slow adoption of autonomous patient communication systems.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially hands-on clinical care like midwifery, has been slower to adopt AI-driven patient communication tools compared to information/finance sectors, with pilots more common than deployment.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by drafting educational materials, generating visual explanations, or offering multilingual translations that a midwife then delivers or refines, improving patient communication efficiency. However, the human midwife remains essential for clinical judgment and trust-building.
Augmentation potentialclaude-sonnet-54/5AI can help midwives prepare clear explanatory materials, translate medical jargon, and support multilingual communication, meaningfully aiding but not replacing the human explanation process.
Task automatabilityclaude-haiku-4-5-202510011/5Explaining medical procedures requires contextual judgment, empathy, and real-time responsiveness to individual patient concerns and literacy levels. Current AI cannot reliably perform this interpersonal communication task end-to-end in clinical settings, and even as a partial assistant it cannot replace the human midwife's clinical authority and emotional presence.
Task automatabilityclaude-sonnet-52/5AI can generate explanatory text but cannot deliver in-person, context-sensitive explanations during labor/delivery care, adjust to real-time patient distress, or answer follow-up clinical questions with accountability, so the bulk of this task cannot be done end-to-end by AI today.'
Adoption barriersclaude-haiku-4-5-202510015/5Midwives have professional and often legal responsibility to personally explain procedures and obtain informed consent; patients expect and prefer direct human contact for sensitive prenatal and delivery counseling. Liability, regulatory requirements around informed consent, and the human-contact requirement create hard barriers to full automation.
Adoption barriersclaude-sonnet-54/5Informed consent and clinical explanation of procedures generally require a licensed provider due to liability, regulatory, and clinical judgment requirements, creating a strong barrier to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Developing, integrating, and maintaining AI systems for patient communication, plus oversight to ensure medical accuracy and appropriateness, remains costlier than having a trained midwife provide explanations as part of their standard workflow.
Cost vs. human wageclaude-sonnet-52/5While generating generic explanatory content is cheap, the clinical, personalized, real-time explanation component still requires a licensed provider, so overall cost savings versus the human midwife's time are limited.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI systems can generate boilerplate procedural explanations or draft patient education materials, no deployed product reliably replaces a midwife's explanations in real clinical encounters. AI chatbots exist but are rarely deployed for actual patient counseling in obstetric settings due to liability and trust concerns.
Technical feasibility todayclaude-sonnet-52/5Patient education chatbots and AI-generated handouts exist but are not deployed as the primary means of explaining procedures in midwifery care; live explanation remains a clinician-delivered function.

Instruct student nurse midwives, medical students, or residents on the birthing process.

16

CI 7–25 · exposure 17 · augmentation 63 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Medical education sectors are adopting AI-assisted tools (simulations, content generation) but remain conservative about replacing human instructors due to accreditation, licensure, and quality-assurance requirements. Adoption remains slow and supplementary.
Sector adoption velocityclaude-sonnet-52/5Healthcare clinical education is a slower-adopting sector for AI-driven instruction due to accreditation requirements, hands-on skill verification, and patient safety concerns, though simulation tech adoption is growing.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can augment instruction by generating practice scenarios, providing background content, and offering supplementary explanations, but the core mentoring and clinical modeling role requires sustained human presence and judgment.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully augment instruction via case simulations, quizzing, personalized study plans, and video-based feedback, enhancing but not replacing the hands-on mentor role.
Task automatabilityclaude-haiku-4-5-202510011/5Instructing learners on the birthing process requires real-time demonstration, adaptive feedback to individual learner needs, and embodied clinical judgment that current AI systems cannot provide end-to-end. While AI can generate educational content, it cannot replace the experiential, hands-on mentorship and clinical modeling that defines effective clinical education.
Task automatabilityclaude-sonnet-52/5AI can supply didactic content and simulations, but hands-on clinical teaching, live supervision during deliveries, and modeling procedural judgment cannot be end-to-end automated with equal quality today.'
Adoption barriersclaude-haiku-4-5-202510015/5Medical education is heavily regulated; instructors must be licensed clinicians, and accrediting bodies mandate human oversight of student competency assessment. Liability and patient safety concerns create hard legal barriers to AI-only instruction in this domain.
Adoption barriersclaude-sonnet-54/5Clinical teaching of birthing procedures requires licensed, credentialed supervisors with legal responsibility for patient safety and trainee competency sign-off, creating strong regulatory and liability barriers.
Cost vs. human wageclaude-haiku-4-5-202510011/5The loaded cost of an experienced nurse midwife instructor is offset by the high cost of implementing reliable AI instruction systems that must integrate with simulation labs, clinical settings, and institutional accreditation requirements, plus ongoing human oversight.
Cost vs. human wageclaude-sonnet-52/5AI tools can cheaply supplement lecture content, but the clinical supervision component still requires a paid, licensed instructor, keeping overall cost comparable to or only modestly less than human-led instruction.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI can produce educational materials and answer questions about birthing processes, but no deployed product reliably delivers the full instructional role—real-time clinical guidance, individualized feedback, assessment of competency, and modeling of clinical decision-making under pressure—that this task requires.
Technical feasibility todayclaude-sonnet-52/5Medical education products (simulation software, AI tutoring, question banks) exist but are supplementary; no deployed product independently instructs learners through actual clinical birthing scenarios at scale.

Establish practice guidelines for specialty areas such as primary health care of women, care of the childbearing family, and newborn care.

7

CI 0–15 · exposure 8 · augmentation 63 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare organizations and professional bodies adopt AI cautiously for guideline development; the task remains firmly in human expert domain with minimal AI displacement in production settings due to regulatory and liability constraints.
Sector adoption velocityclaude-sonnet-52/5Healthcare guideline development is a slow, institutionally conservative process with limited AI integration into formal guideline-setting bodies despite AI's broader growth in clinical documentation tools.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist midwives with literature retrieval, evidence synthesis summaries, and initial draft organization, moderately raising their productivity in the research phase; however, the core judgment and authority remain with the human expert.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist midwives by synthesizing research literature, summarizing evidence, and drafting guideline language, substantially speeding up the preparatory work even though final judgment remains human.
Task automatabilityclaude-haiku-4-5-202510011/5Establishing clinical practice guidelines requires deep domain expertise, synthesis of research evidence, understanding of organizational context, legal liability considerations, and professional judgment that current AI cannot perform end-to-end. While AI can assist with literature review or drafting, a credentialed midwife must lead and take responsibility for guidelines.
Task automatabilityclaude-sonnet-52/5Drafting clinical guideline text can be partially assisted by AI, but establishing authoritative practice guidelines requires clinical judgment, consensus-building, and accountability that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Establishing practice guidelines for clinical care is a regulated, high-liability activity requiring licensed healthcare professionals to author and sign off. Medical and midwifery board oversight, accreditation bodies, and patient safety liability create hard barriers to full automation.
Adoption barriersclaude-sonnet-55/5Establishing clinical practice guidelines requires licensed, credentialed practitioners and professional bodies to author and approve standards, with significant liability implications, making this a hard-barrier task.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of AI-generated guideline drafting plus required expert human review, validation, and legal oversight would exceed the cost of experienced midwives directly authoring guidelines, as the human review burden remains substantial.
Cost vs. human wageclaude-sonnet-52/5While AI drafting tools are cheap, the human review, clinical validation, and liability sign-off required keep overall costs comparable to or only modestly below fully human-led guideline development.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can autonomously create clinically valid, legally defensible practice guidelines for midwifery care. This task requires live clinical expertise, regulatory knowledge, and institutional accountability that exceeds current AI system capabilities.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously establishes clinical practice guidelines in production; this remains a human expert-committee process, with AI at most a research or drafting aid.

Consult with or refer patients to appropriate specialists when conditions exceed the scope of practice or expertise.

7

CI 3–11 · exposure 5 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of clinical decision AI remains cautious and pilot-heavy; referral routing in particular has seen limited production deployment due to risk aversion and regulatory uncertainty.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially maternal care requiring physical exams and licensed judgment, adopts AI more slowly than digital-first sectors, and this specific clinical judgment task sees minimal AI deployment currently.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by prompting consideration of specialists, surfacing guidelines, and organizing patient history—useful scaffolding—but the clinical judgment call itself remains fundamentally human-centered and legally non-delegable.
Augmentation potentialclaude-sonnet-53/5AI can assist by flagging risk factors from clinical data, suggesting relevant specialists, or drafting referral documentation, but the core judgment and communication remain human-driven.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires clinical judgment about patient complexity, diagnostic awareness, and knowledge of when a condition falls outside midwifery scope—core medical reasoning that current AI systems cannot reliably perform end-to-end in real patient contexts.
Task automatabilityclaude-sonnet-51/5This task requires clinical judgment about when a patient's condition exceeds one's scope of practice, plus interpersonal coordination with specialists—AI cannot independently make this judgment call or execute the referral relationship.
Adoption barriersclaude-haiku-4-5-202510015/5Scope-of-practice determination is legally and professionally mandated for licensed midwives; accountability for missed referrals and adverse outcomes falls on the clinician, creating hard liability barriers to automation.
Adoption barriersclaude-sonnet-55/5This is a core, legally-mandated professional responsibility—midwives are licensed practitioners who must recognize and act on scope-of-practice limits, and referral decisions carry direct liability and patient safety implications.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-assisted flagging of referral candidates has emerged, but integration into clinical workflows and the need for human oversight mean costs approach or match the midwife's time for this judgment-heavy task.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this end-to-end, so cost comparison favors the human entirely; any AI involvement is only a minor administrative aid, not a replacement of the judgment/referral function.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can assist in surfacing relevant specialist options or flagging red-flag symptoms, no deployed product reliably makes the clinical decision to refer in production; the liability and stakes of incorrect triage remain too high for autonomous performance.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously identifies scope-of-practice boundaries and initiates specialist referrals for midwifery care; this remains a clinical decision-making process requiring licensed practitioners.

Initiate emergency interventions to stabilize patients.

5

CI 3–7 · exposure 5 · augmentation 50 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of AI is cautious in acute-care settings; emergency stabilization remains one of the most heavily guarded human domains. While monitoring and alert systems are deployed, actual intervention initiation is not being automated at scale in production.
Sector adoption velocityclaude-sonnet-52/5Healthcare broadly is a slower-adopting sector for autonomous action, and emergency/acute care specifically has seen minimal deployment of AI performing physical interventions.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist via real-time clinical decision support, vital-sign analysis, and protocol reminders during stabilization, improving clinician confidence and speed. However, the core interventional task remains human-directed, limiting augmentation to supporting layers rather than transforming primary performance.
Augmentation potentialclaude-sonnet-53/5AI-based monitoring, early warning scoring, and decision-support systems can help flag deterioration and suggest protocols, aiding midwives' situational awareness during emergencies.
Task automatabilityclaude-haiku-4-5-202510011/5Emergency stabilization requires real-time clinical judgment, physical intervention (medications, airway management, resuscitation), and immediate adaptation to patient response. Current AI cannot physically act, make split-second life-or-death decisions, or reliably handle the full dynamic complexity of acute emergencies.
Task automatabilityclaude-sonnet-51/5Emergency stabilization requires real-time physical intervention, hands-on procedures, and split-second judgment under uncertainty that no current AI system can perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Emergency interventions are legally and clinically required to be performed or directly overseen by a licensed nurse midwife. Liability, licensure requirements, and regulatory frameworks create hard barriers to full automation or substitution.
Adoption barriersclaude-sonnet-55/5Emergency medical intervention is tightly regulated, requires licensed clinical personnel, and carries severe liability and life-safety consequences that legally mandate human performance.
Cost vs. human wageclaude-haiku-4-5-202510011/5Emergency stabilization is a high-acuity clinical task requiring licensed human presence. The cost of AI infrastructure, oversight, and liability to catch errors in life-or-death situations far exceeds the wage of a specialized midwife providing direct care.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so cost comparison favors the human entirely; any AI role is purely advisory and adds cost rather than replacing labor.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can assist with triage protocols and decision support, no deployed system reliably performs emergency stabilization end-to-end. Existing AI decision-support tools have material limitations in novel or atypical presentations and cannot execute physical interventions; they support rather than replace clinician judgment.
Technical feasibility todayclaude-sonnet-51/5No deployed product initiates emergency medical interventions autonomously; clinical decision support tools exist but do not perform the physical act of stabilizing a patient.

Monitor fetal development by listening to fetal heartbeat, taking external uterine measurements, identifying fetal position, or estimating fetal size and weight.

4

CI 0–7 · exposure 5 · augmentation 63 · importance 5.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare adoption of AI is growing in imaging and data analysis, but direct automation of hands-on fetal monitoring remains rare in production; most deployments remain pilots or adjuncts to standard clinical practice rather than replacements.
Sector adoption velocityclaude-sonnet-51/5Healthcare physical examination tasks, especially in obstetrics, show very slow AI adoption due to the inherently physical, hands-on nature of the work and stringent regulatory/safety requirements.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist by analyzing CTG strips, ultrasound images, or measurements captured by the clinician, flagging anomalies and supporting rapid decision-making, while the midwife retains responsibility for examination, judgment, and patient safety.
Augmentation potentialclaude-sonnet-53/5AI-enabled fetal monitoring devices (e.g., algorithmic interpretation of fetal heart rate patterns, ultrasound-based estimation tools) can assist in interpreting data and flagging anomalies, augmenting the midwife's assessment even though the physical exam remains human-performed.
Task automatabilityclaude-haiku-4-5-202510011/5This task fundamentally requires real-time physical examination (auscultation, palpation, measurement) and clinical judgment that current AI cannot perform end-to-end without a human clinician present and performing the physical acts.
Task automatabilityclaude-sonnet-51/5This requires hands-on physical examination, palpation, and use of a fetoscope/doppler directly on a patient's body, which current AI systems cannot perform end-to-end without robotic embodiment that doesn't exist in clinical practice.
Adoption barriersclaude-haiku-4-5-202510015/5A licensed midwife or physician must legally perform fetal assessment and clinical monitoring; liability for misinterpretation of fetal status is high, and the task involves direct patient contact and legal/regulatory requirements for human licensure and accountability.
Adoption barriersclaude-sonnet-55/5This is a licensed clinical procedure requiring hands-on physical examination by a credentialed midwife or physician, with significant liability implications for maternal-fetal health outcomes.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of AI analysis of captured data is marginal compared to the loaded wage of a nurse midwife, and it does not replace the core clinical labor of examination and decision-making, making substitution economically unfavorable.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this physical task, so cost comparison favors the human provider entirely since AI cannot replace the physical examination component.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can assist with analysis of CTG tracings or ultrasound images post-capture, no deployed system can independently perform the complete physical assessment—listening to fetal heartbeat, taking measurements, identifying position—without trained human clinicians doing the hands-on work.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously performs physical fetal monitoring exams like Leopold's maneuvers or manual fetal heartbeat auscultation; this remains a hands-on clinical skill performed by trained providers.

Manage newborn care during the first weeks of life.

3

CI 0–5 · exposure 5 · augmentation 50 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare remains a laggard sector for full-task automation of direct patient care, especially for vulnerable populations like newborns. Adoption is limited to narrow monitoring aids rather than care delegation.
Sector adoption velocityclaude-sonnet-51/5Healthcare delivery for hands-on newborn care is a low-digitization, physically-mediated sector with minimal AI displacement in production.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist with trend analysis of vital signs, feeding logs, and clinical documentation, helping midwives organize information and flag deviations—useful but not transformative to the core clinical skills required.
Augmentation potentialclaude-sonnet-53/5AI can assist with charting, decision support, and educational materials for parents, but doesn't transform the hands-on care process itself.
Task automatabilityclaude-haiku-4-5-202510011/5Newborn care requires continuous physical examination, hands-on intervention (feeding, diaper changes, positioning), vital sign monitoring, and rapid clinical decision-making in response to patient state. Current AI cannot perform these embodied, safety-critical tasks end-to-end.
Task automatabilityclaude-sonnet-51/5This involves hands-on physical assessment, feeding support, and clinical judgment on a live infant that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Midwifery and neonatal care are heavily regulated; legal and professional standards require a licensed clinician to directly perform and sign off on newborn assessment and care. Liability for adverse outcomes falls on licensed practitioners.
Adoption barriersclaude-sonnet-55/5Newborn care requires licensed clinical personnel, involves significant liability, and is heavily regulated, creating hard barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Newborn care demands continuous human presence for safety, responsiveness, and liability reasons. The cost of an AI system plus required human supervision would exceed the cost of direct human care.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical, clinical labor involved, so no meaningful cost comparison favors AI for the core task.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI can support documentation and monitoring alerts, but no deployed system reliably performs the full scope of newborn physical care, feeding assessment, or emergency response that defines this task in practice.
Technical feasibility todayclaude-sonnet-51/5No deployed product provides autonomous newborn clinical management; at most AI supports documentation or triage information, not hands-on care.

Provide prenatal, intrapartum, postpartum, or newborn care to patients.

1

CI 0–3 · exposure 0 · augmentation 50 · importance 5.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare adoption of AI remains focused on diagnostic support and administrative tasks, not direct clinical care delivery. Midwifery is a relationship-intensive, hands-on specialty with no meaningful sector-wide AI displacement of core clinical functions.
Sector adoption velocityclaude-sonnet-52/5Healthcare delivery, especially hands-on maternal care, adopts AI slowly due to regulatory, safety, and licensure constraints, with only administrative or diagnostic support tools seeing uptake.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist midwives through monitoring dashboards, risk flagging from patient data, documentation support, and clinical decision aids during prenatal and postpartum review. However, the intrapartum phase—the heart of midwifery—offers limited augmentation opportunities for a human in active labor management.
Augmentation potentialclaude-sonnet-53/5AI can assist with charting, risk-flagging from vitals/labs, and patient education materials, but does not touch the core physical care tasks.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires direct physical examination, hands-on intervention during labor/delivery, real-time clinical decision-making under variable conditions, and continuous human presence. AI cannot perform physical assessments, delivery assistance, or emergency interventions that define midwifery care.
Task automatabilityclaude-sonnet-51/5This task requires hands-on physical examination, delivery assistance, and real-time clinical judgment during birth that no current AI system can perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Midwifery is a licensed profession; regulations in virtually all jurisdictions require a credentialed midwife to provide direct prenatal, intrapartum, and postpartum care. Legal and liability requirements create hard barriers to automation.
Adoption barriersclaude-sonnet-55/5Nurse midwifery is a licensed clinical profession with strict scope-of-practice laws, liability requirements, and mandatory human presence for childbirth and newborn care.
Cost vs. human wageclaude-haiku-4-5-202510011/5Even with future AI advances, the infrastructure, liability, and oversight required to partially automate aspects of midwifery care would far exceed the cost of a human midwife for decades. The task involves irreducible human judgment and presence.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical and clinical labor involved, so there is no viable cost comparison—human midwives remain necessary for all direct care delivery.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system can independently provide prenatal exams, manage labor progression, deliver a baby, or assess newborn health. Clinical decision support exists but does not meet the bar of performing the core task end-to-end.
Technical feasibility todayclaude-sonnet-51/5No deployed product provides prenatal, intrapartum, postpartum, or newborn care; AI is at most used for documentation or decision-support adjuncts, not the care itself.

Prescribe medications as permitted by state regulations.

1

CI 0–3 · exposure 0 · augmentation 63 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare sectors are conservative on autonomous prescribing due to regulatory and safety constraints; there is no meaningful adoption of AI-driven prescription issuance even in the most digitized healthcare organizations.
Sector adoption velocityclaude-sonnet-52/5Healthcare prescribing workflows adopt AI mainly for decision support and documentation, but actual prescribing authority automation is essentially absent due to regulatory constraints.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist nurse midwives by suggesting appropriate medications, flagging drug interactions, and providing evidence-based guidance, but the clinician retains full responsibility for the prescriptive decision.
Augmentation potentialclaude-sonnet-54/5AI clinical decision-support and drug-interaction tools can meaningfully assist midwives in selecting appropriate medications and checking safety, improving efficiency while the human retains prescribing authority.
Task automatabilityclaude-haiku-4-5-202510011/5Medication prescribing requires human judgment about patient history, contraindications, allergies, and clinical context that AI cannot reliably assess end-to-end; additionally, prescriptive authority is a licensed clinical function that cannot be delegated to AI systems.
Task automatabilityclaude-sonnet-51/5Prescribing requires clinical judgment, physical assessment, and legal accountability that current AI cannot autonomously perform end-to-end; no time-saving substitution meets the equal-quality bar.
Adoption barriersclaude-haiku-4-5-202510015/5Prescriptive authority is protected by state medical boards, licensure requirements, and liability law; only licensed providers can legally issue prescriptions, creating an absolute legal barrier to automation.
Adoption barriersclaude-sonnet-55/5Prescribing is tightly regulated and requires a licensed, credentialed provider (nurse midwife with prescriptive authority) to legally authorize medications, creating a hard legal barrier to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The task cannot be automated, so cost comparison is not applicable; a nurse midwife's time is required regardless.
Cost vs. human wageclaude-sonnet-51/5AI cannot legally or practically substitute for the prescriber, so there is no valid AI-only cost comparison; the human cost remains mandatory.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system can independently prescribe medications—this remains a human clinician responsibility in all production healthcare settings, though AI may assist in drug selection recommendations.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously prescribes medications for patients; clinical decision-support tools exist but require licensed provider sign-off, so this remains research/assistive stage only.

Provide primary health care, including pregnancy and childbirth, to women.

1

CI 0–3 · exposure 0 · augmentation 50 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Midwifery remains a human-centered, physically present clinical role with minimal AI displacement. Adoption of AI assistive tools in maternity care is slow and limited to data-analytic or administrative functions; there is no production evidence of deep or fast AI adoption that displaces midwife labor.
Sector adoption velocityclaude-sonnet-52/5Healthcare delivery, especially hands-on obstetric care, is a slow-adopting sector due to regulation, safety concerns, and physical necessity of care.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist midwives through decision-support tools (risk flagging in electronic health records), fetal monitoring analysis, and automated documentation, improving efficiency in some subtasks. However, augmentation is limited because the core work—patient interaction, clinical judgment, hands-on delivery care—remains fundamentally human-centered.
Augmentation potentialclaude-sonnet-53/5AI can assist with documentation, risk flagging, decision support, and patient education, improving efficiency without replacing hands-on care.
Task automatabilityclaude-haiku-4-5-202510011/5This task fundamentally requires hands-on clinical assessment, physical examination, decision-making in real-time clinical contexts, and direct patient care during labor and delivery—functions that current AI cannot perform end-to-end. No existing AI system can replace the core competencies of pregnancy monitoring, childbirth management, and primary health care delivery.
Task automatabilityclaude-sonnet-51/5This task involves hands-on physical examination, delivering babies, and emergency clinical decision-making that cannot be performed end-to-end by AI systems today.
Adoption barriersclaude-haiku-4-5-202510015/5Licensure, legal responsibility, and liability requirements mandate that a licensed nurse midwife (or physician) must personally manage pregnancy, childbirth, and complications. Regulatory frameworks (e.g., state nursing boards, medical licensing) create hard legal barriers that explicitly require human professional judgment and accountability.
Adoption barriersclaude-sonnet-55/5Nurse midwifery requires licensure, clinical certification, and legal authority to perform medical procedures and prescribe care, with high liability for errors in maternal/infant health.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI tools that support specific aspects of midwifery (imaging analysis, documentation) are far cheaper than nurse midwife salaries, but they cannot substitute for the human clinician; the comparison is not meaningful because the task cannot be automated. Integration and oversight costs would add to AI expenses without replacing the need for a midwife.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical care delivery, so the relevant cost comparison is not applicable—human labor remains required at full cost.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs this task as a whole; AI systems exist only for narrow subtasks (e.g., fetal monitoring interpretation, administrative support). The full scope of midwifery care—diagnosis, intervention, complications management, and patient counseling—remains outside the capability of any production AI system.
Technical feasibility todayclaude-sonnet-51/5No deployed product provides primary care or delivers babies autonomously; AI is at most a decision-support or documentation tool alongside a human clinician.

Provide patients with direct family planning services, such as inserting intrauterine devices, dispensing oral contraceptives, and fitting cervical barriers, including cervical caps or diaphragms.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This task involves direct clinical care and medical device insertion that cannot be delegated to non-human agents under current law and regulation. Adoption velocity of AI for this specific task is zero, as it remains firmly in the domain of licensed practitioners.
Sector adoption velocityclaude-sonnet-51/5Healthcare's hands-on clinical procedures remain among the least AI-penetrated tasks, with physical intervention requirements creating a hard adoption floor.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist marginally with educational content, patient intake documentation, or contraceptive option comparison tools, but the core clinical procedures and patient fitting require direct human expertise and cannot be meaningfully augmented by current AI.
Augmentation potentialclaude-sonnet-52/5AI may help with patient education, documentation, or decision support around contraceptive choice, but offers minimal assistance for the actual physical insertion or fitting procedure.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires hands-on clinical procedures (IUD insertion, physical fitting of cervical barriers) and direct patient contact that cannot be performed by current AI systems. No meaningful automation of the core procedural and physical components exists.
Task automatabilityclaude-sonnet-51/5This requires direct physical manipulation and insertion of medical devices into a patient's body, which current AI systems cannot perform as they lack physical embodiment for clinical procedures.
Adoption barriersclaude-haiku-4-5-202510015/5Substantial legal and regulatory barriers exist: only licensed healthcare providers (nurse midwives, physicians) are authorized to perform these clinical procedures, insert medical devices, and prescribe contraceptives. Legal liability and scope-of-practice statutes create hard barriers to any substitution.
Adoption barriersclaude-sonnet-55/5This is a licensed clinical procedure requiring a credentialed nurse midwife or physician, with direct physical patient contact, informed consent, and legal scope-of-practice requirements.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI cannot perform this task at all, making cost comparison moot. The loaded cost of a nurse midwife performing these services is considerably lower than any hypothetical system architecture that might one day assist, given the specialization required.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical task at all, so there is no viable cost comparison—the human provider remains the only option.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs intrauterine device insertion, contraceptive dispensing, or fitting of cervical barriers. These are inherently procedural tasks requiring licensed human clinicians and direct physical intervention that is not automatable by current technology.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical gynecological procedures like IUD insertion or diaphragm fitting; this remains entirely in the domain of human clinicians.

Perform physical examinations by taking vital signs, checking neurological reflexes, examining breasts, or performing pelvic examinations.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI for performing hands-on physical exams is not occurring because it is not technologically feasible. Midwifery practice remains deeply grounded in in-person clinical assessment.
Sector adoption velocityclaude-sonnet-51/5Healthcare physical examination remains a highly hands-on domain with minimal AI penetration for the actual physical act, despite AI use in EHR documentation or decision support surrounding it.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist modestly by automating documentation of vital signs or providing decision-support for interpretation, but the core examination task—the tactile and observational assessment itself—remains entirely human-driven with minimal augmentation benefit.
Augmentation potentialclaude-sonnet-52/5AI can assist with documentation, decision support, or interpreting vital sign trends, but offers no meaningful assistance to the physical act of examining a patient's body.
Task automatabilityclaude-haiku-4-5-202510011/5Physical examinations require direct tactile assessment, palpation, and real-time clinical judgment that current AI cannot perform. While vital signs can be monitored by devices, the full examination—reflexes, breast exams, pelvic exams—requires human hands-on interaction that AI systems cannot execute today.
Task automatabilityclaude-sonnet-51/5This requires direct physical manipulation, touch-based assessment, and real-time clinical judgment during hands-on examination; no current AI system can perform physical exams autonomously.
Adoption barriersclaude-haiku-4-5-202510015/5Medical examination requires direct contact with patients and involves liability, legal accountability, and patient dignity—nurse midwives must personally perform and sign off on findings. Regulatory and licensure frameworks mandate that a qualified human conduct physical examinations.
Adoption barriersclaude-sonnet-55/5Licensed clinical practitioners are legally required to perform physical examinations, and there is no regulatory or feasible pathway for non-human automation of hands-on patient care in this manner.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI cannot perform this task at all, making cost comparison inapplicable. When an AI system cannot execute the core work, it remains more expensive than the human alternative.
Cost vs. human wageclaude-sonnet-51/5AI cannot perform this physical task at all, so there is no viable cost comparison—the human is the only option, making AI infinitely more 'expensive' in the sense of being non-functional.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can reliably perform the physical contact and tactile assessment required for these examinations. Clinical documentation and image analysis exist, but actual hands-on examination remains entirely in the human domain.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical examinations like pelvic exams or reflex testing; this remains entirely outside the scope of current AI products, which are digital/software-based.

Related occupations — Healthcare Practitioners & Technical

How to read this

A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.

What would change this score

New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.