Obstetricians and Gynecologists
29-1218.00Median wage $292,910/yr21,260 employed (US)Rank #709 of 923 scored · top 77% by substitution
Provide medical care related to pregnancy or childbirth. Diagnose, treat, and help prevent diseases of women, particularly those affecting the reproductive system. May also provide general care to women. May perform both medical and gynecological surgery functions.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution19
Exposure20
Augmentation64
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
15 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%18
panel mean rating 1.7/5 → substitution pressure 18/100
Technical feasibility todayw 20%23
panel mean rating 1.9/5 → substitution pressure 23/100
Cost vs. human wagew 15%22
panel mean rating 1.9/5 → substitution pressure 22/100
Adoption barriersw 20%inverted — strong barriers lower the score10
panel mean rating 4.6/5 (barrier strength) → substitution pressure 10/100
Sector adoption velocityw 10%24
panel mean rating 2.0/5 → substitution pressure 24/100
Task breakdown (15 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.
Prepare government and organizational reports on birth, death, and disease statistics, workforce evaluations, or the medical status of individuals.
67CI 62–71 · exposure 70 · augmentation 88 · importance 3.4/5 · click for rater detail
Prepare government and organizational reports on birth, death, and disease statistics, workforce evaluations, or the medical status of individuals.
67| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Mid-to-large healthcare systems and health departments have adopted analytics and reporting tools, but many smaller practices and legacy organizations still rely on manual compilation; adoption is uneven and driven by regulatory/accreditation pressure rather than market demand. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare administrative and reporting functions are seeing growing but uneven AI adoption, with hospital systems piloting automated analytics while many still rely on manual compilation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tools that auto-populate fields, flag anomalies, and generate draft narratives substantially boost physician productivity in preparing accurate, timely reports without removing physician accountability. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI tools can significantly speed up drafting, aggregating, and formatting statistical reports, letting physicians focus on review and interpretation rather than manual compilation. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can extract data from medical records, compile statistics, and generate structured reports with high accuracy and significant time savings. However, interpretation of nuanced medical status assessments and contextual judgment on sensitive health data still typically require physician review and sign-off. |
| Task automatability | claude-sonnet-5 | 4/5 | Compiling statistical reports on births, deaths, disease incidence and workforce evaluations is largely structured data aggregation and narrative summarization, which current AI/LLM tools handle well when given structured inputs.rror |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Government reporting requirements and compliance (HIPAA, state health department rules) create mandatory oversight and physician attestation in many jurisdictions; however, these are procedural rather than legal prohibitions on automation itself, allowing AI-assisted workflows. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Government reporting often has regulatory formatting and certification requirements, and a licensed physician may need to attest to medical status data, creating moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated data extraction and report generation cost cents to dollars per report, compared to physician time at $100+ per hour; AI is one to two orders of magnitude cheaper once infrastructure is in place. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated report generation from existing EHR/registry data is far cheaper than physician time spent compiling and writing these reports, though some integration and review cost remains. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed systems (EHR-integrated analytics, report-generation tools, and data extraction pipelines) routinely produce birth/death statistics and workforce reports in clinical and public health settings. Accuracy is generally high, though integration with legacy systems and validation workflows remain common. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | EHR and analytics platforms already auto-generate many statistical/compliance reports, but customization for specific government/organizational formats and validation still requires human review, limiting full production reliability. |
Collect, record, and maintain patient information, such as medical histories, reports, or examination results.
62CI 50–74 · exposure 62 · augmentation 88 · importance 4.8/5 · click for rater detail
Collect, record, and maintain patient information, such as medical histories, reports, or examination results.
62| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Healthcare, particularly large obstetric practices and hospital systems, shows strong adoption of AI documentation tools and ambient voice transcription; deployment in production is measurable and accelerating, though smaller practices lag. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare is adopting AI documentation tools moderately fast but overall sector digitization and regulatory caution keep it behind fast-moving sectors like finance or software. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI transforms physician and staff productivity by auto-generating structured clinical notes, flagging missing data, and pre-populating forms, allowing the human clinician to focus on patient care and clinical review rather than data entry. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI scribes and summarization tools meaningfully reduce charting time and can pre-populate structured fields, materially boosting physician productivity while they remain in the loop for verification. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI systems can reliably extract, structure, and record clinical data from patient intake forms, electronic health records, and examination notes with high accuracy, achieving significant time savings. However, nuanced clinical judgment about what to record and when to flag information for physician review still benefits from human oversight, preventing a full 5. |
| Task automatability | claude-sonnet-5 | 3/5 | AI ambient scribes and NLP tools can transcribe, summarize, and structure patient history and exam findings into EHR fields, but clinicians still must verify accuracy and enter/confirm clinical data, so only partial time savings are realized end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | HIPAA compliance, data privacy regulations, and institutional EHR governance impose oversight requirements and integration friction. However, there is no legal barrier preventing AI from performing this task; medical personnel may prefer direct patient contact, but the task itself is not gatekept. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Physicians remain legally responsible for accuracy of medical records and must review/sign documentation, creating moderate liability-driven barriers even though scribing assistance itself is not restricted by licensure. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-driven documentation automation costs a fraction of clinical staff labor—inference, integration, and minimal oversight are orders of magnitude cheaper than employing medical records personnel or physician time spent on manual charting. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Subscription costs for ambient scribe tools are modest compared to physician time saved, but integration, EHR customization, and required human review keep the net savings moderate rather than an order-of-magnitude reduction. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature EHR systems with AI-powered data extraction, clinical documentation assistants (e.g., ambient voice-to-text with auto-structuring), and medical information extraction models are deployed at scale in obstetric and gynecologic practices today. Error rates are low for routine documentation, though complex cases may require human review. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Ambient documentation products (e.g., DAX Copilot, Nuance, Abridge) are deployed in real clinical settings and reduce documentation burden, but adoption is uneven and error correction/oversight is still required, limiting full reliability. |
Direct and coordinate activities of nurses, students, assistants, specialists, therapists, and other medical staff.
30CI 3–57 · exposure 33 · augmentation 63 · importance 4.2/5 · click for rater detail
Direct and coordinate activities of nurses, students, assistants, specialists, therapists, and other medical staff.
30| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Healthcare institutions are adopting workflow management tools and electronic scheduling, but adoption remains uneven and primarily pilot-stage for autonomous coordination. Most large hospitals have some system in place, but genuine AI-driven delegation is still emerging. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administrative/coordination workflows are adopting AI slowly for scheduling and documentation, but leadership/direction of clinical teams remains largely untouched by automation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI dashboards, predictive alerts, and scheduling suggestions can substantially improve a physician's ability to coordinate staff in real time, freeing them from administrative burden while they retain decision authority. This augmentation model is already proving valuable in some healthcare settings. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with scheduling, task tracking, and communication support among team members, improving efficiency, but does not replace the human-led coordination and decision-making. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI agents can handle scheduling, task assignment, resource allocation, and monitoring of team workflows with significant automation potential. However, real-time interpersonal conflict resolution and nuanced mentoring of clinical staff still require human judgment, preventing a full 5 rating. |
| Task automatability | claude-sonnet-5 | 1/5 | Directing and coordinating a live medical team requires real-time judgment, authority, interpersonal leadership, and physical presence that current AI cannot replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Clinical governance and liability frameworks in many jurisdictions require a licensed physician to retain ultimate authority over staff direction and task assignment in obstetrics/gynecology. Physicians are reluctant to delegate real-time clinical coordination fully to AI due to patient safety accountability. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Clinical supervision, delegation of care duties, and staff direction are governed by licensure, hospital credentialing, and legal responsibility that require a physician to perform this role. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Once implemented, AI scheduling and task-coordination systems have negligible per-use cost compared to physician time spent on administrative coordination. A physician's loaded wage is substantially higher than the inference and integration cost of scheduling tools. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this managerial task, so any comparison favors the human physician who must retain accountability and leadership. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Workflow management and scheduling software exists and sees some deployment in healthcare systems, but most implementations are still narrow (scheduling only) or hybrid with heavy manual oversight. Few organizations have fully autonomous AI directing clinical team coordination. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product manages or directs clinical staff teams; scheduling/communication tools exist but do not perform the supervisory/coordination role itself. |
Advise patients and community members concerning diet, activity, hygiene, and disease prevention.
27CI 25–29 · exposure 25 · augmentation 75 · importance 4.1/5 · click for rater detail
Advise patients and community members concerning diet, activity, hygiene, and disease prevention.
27| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare remains cautious and heavily regulated; while digital health tools are spreading, direct replacement of physician counseling by autonomous AI remains rare in clinical practice due to liability, licensure, and patient-trust norms. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, particularly clinical patient counseling, adopts AI tools cautiously due to regulatory, liability, and trust concerns, with pilots more common than widespread production use for this specific task. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can effectively assist by retrieving evidence-based guidelines, suggesting talking points on diet and prevention, and generating patient-friendly educational materials that the physician then delivers or refines, materially enhancing counseling productivity. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can efficiently draft patient education materials, personalize handouts, and support pre-visit information gathering, meaningfully augmenting physician efficiency while the physician retains the counseling role. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate general health advice on diet, activity, and hygiene, the task requires personalized clinical judgment tailored to individual patient histories, risk factors, and pregnancy/gynecological status—factors that current systems cannot reliably assess or contextualize without substantial human input and review. |
| Task automatability | claude-sonnet-5 | 2/5 | General health advice generation is possible with current AI, but personalized advising requires patient context, examination findings, and relationship-based counseling that current systems cannot fully replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | OB/GYNs are licensed physicians whose medical advice carries legal and liability responsibility; regulatory frameworks and medical ethics require a licensed clinician to provide health counsel directly to patients, creating substantial barriers to unsupervised automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Medical advice giving is tied to licensure and liability; physicians are expected to personally counsel patients, especially in areas with legal and health-outcome sensitivity like pregnancy and reproductive health. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The cost of AI systems with appropriate oversight, integration into clinical workflows, and liability safeguards approaches or exceeds the cost of brief provider-patient consultation time, especially when factoring in human supervision for medical accuracy. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-generated educational content is cheap to produce, but integration into clinical workflow with physician oversight for accuracy and liability narrows the cost advantage over the physician's time actually spent. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some AI chatbots and decision-support tools can deliver standardized health information, but deployed products lack the clinical integration, liability handling, and ability to address patient-specific concerns that this medical advisory task demands in a real clinical setting. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Chatbots and patient portals offer generic health education content, but no deployed product reliably substitutes for a physician's personalized counseling on diet, activity, hygiene, and disease prevention in obstetric/gynecologic care. |
Analyze records, reports, test results, or examination information to diagnose medical condition of patient.
20CI 20–20 · exposure 25 · augmentation 63 · importance 4.9/5 · click for rater detail
Analyze records, reports, test results, or examination information to diagnose medical condition of patient.
20| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | OB/GYN practices have adopted some imaging and lab software aids, but diagnostic automation remains limited due to regulatory stringency, malpractice liability, and the high stakes of maternal and gynecologic care. Adoption lags compared to lower-liability administrative or billing tasks. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially clinical diagnostic work, adopts AI cautiously due to regulatory, liability, and workflow integration barriers, with pilots more common than widespread production deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can productively assist by flagging abnormalities in imaging, summarizing lab trends, and organizing patient data for physician review. However, the integration of these aids into clinical workflow is still emerging and varies significantly across practice settings. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools (e.g., decision support, literature synthesis, imaging analysis) meaningfully assist physicians in reviewing test results and records, improving efficiency and catching patterns while the physician retains diagnostic authority. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in analyzing medical imaging and lab results, the task requires integrating complex patient history, physical findings, and contextual factors that demand human judgment. Current AI systems lack reliability for end-to-end diagnosis in obstetric/gynecologic cases with sufficient consistency to meet the 50% time-saving threshold without substantial physician review. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can assist in flagging abnormal results or suggesting differentials but full end-to-end diagnostic synthesis integrating history, exam findings, and clinical judgment for OB/GYN patients is not reliably automatable today with equal quality.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Licensed physicians are legally and professionally required to establish diagnoses and sign off on medical findings; liability for missed or incorrect diagnoses creates hard regulatory barriers. Patient care relationships and malpractice risk mean this task cannot be delegated to an automated system without full physician responsibility. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Diagnosis is a licensed medical act requiring physician judgment and legal accountability; regulatory and liability frameworks mandate a physician sign off on diagnostic conclusions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Diagnostic AI tools and radiography software require substantial licensing, infrastructure, and physician oversight costs that approach or exceed the cost-benefit of a physician's time on routine analysis, especially given liability and verification burdens in this specialty. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Given the need for extensive human oversight, liability review, and integration with EHR systems, AI cost savings are offset by required physician verification, making cost roughly comparable rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some narrow diagnostic AI applications exist (e.g., imaging analysis), but deployed products for comprehensive OB/GYN diagnosis remain limited and typically require significant physician validation. Maternity and gynecological conditions involve high-stakes diagnosis where production systems remain developmental rather than reliable. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some clinical decision support and diagnostic aid tools exist (e.g., imaging analysis, risk calculators) but no deployed product autonomously performs comprehensive diagnostic synthesis for OB/GYN patients in production at scale. |
Monitor patients' conditions and progress and reevaluate treatments as necessary.
16CI 11–20 · exposure 17 · augmentation 63 · importance 4.7/5 · click for rater detail
Monitor patients' conditions and progress and reevaluate treatments as necessary.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare AI adoption remains cautious and heavily regulated. While some hospitals deploy clinical decision support and monitoring alerts, autonomous or near-autonomous treatment reevaluation has seen minimal real-world production adoption due to liability, validation, and regulatory friction. Pilots and research installations vastly outnumber operational deployments. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially high-liability specialties like OB/GYN, adopts AI slowly due to regulatory, safety, and workflow integration challenges despite some EHR-embedded decision support tools. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by surfacing relevant data (prior records, complication flags, lab trends) and providing decision-support prompts, reducing cognitive load on clinicians. However, augmentation is bounded by integration friction and physician skepticism; it enhances workflow but does not yet transform productivity on this task in mainstream practice. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-based monitoring dashboards, risk stratification algorithms, and alerting systems meaningfully help physicians track patient trends and flag concerns, improving efficiency while the physician retains decision authority. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with interpreting some clinical data (lab results, imaging), the core task—real-time clinical judgment about patient condition trajectories and treatment adjustments—requires contextual synthesis, patient rapport, and accountability that current AI cannot perform autonomously. Meaningful automation would require AI to make treatment decisions independently, which neither deployed systems nor practitioners would accept. |
| Task automatability | claude-sonnet-5 | 1/5 | Monitoring OB/GYN patients and dynamically reevaluating treatment requires continuous clinical judgment, physical exam findings, and integration of nuanced patient context that current AI cannot autonomously perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task is legally and professionally gated: only licensed physicians can make clinical judgments about patient management and treatment changes in most jurisdictions. Medical malpractice liability, standard-of-care requirements, and direct patient care mandates create hard barriers to autonomous AI deployment. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Medical licensure, malpractice liability, and legal requirements mandate that a physician personally evaluate and adjust treatment plans, especially in obstetric/gynecologic care with high-stakes outcomes. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Obstetrician-gynecologists earn high salaries ($250k+), and the liability and oversight costs for AI-driven clinical decisions remain substantial. AI infrastructure for patient monitoring is costly relative to its output, and does not yet approach an order-of-magnitude cost advantage over human clinician time on this specific task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI monitoring tools add cost on top of physician oversight rather than replacing it, so overall cost per patient-equivalent is not meaningfully lower than physician-only care. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools exist for specific subtasks (vital sign monitoring alerts, EHR documentation support), but no deployed product reliably performs the full task of clinical monitoring and treatment reevaluation with the judgment quality required in obstetrics/gynecology. Current systems operate in narrow, well-defined domains rather than the adaptive decision-making this role demands. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some clinical decision-support and monitoring tools (e.g., fetal monitoring alerts, risk-scoring algorithms) are deployed, but they assist rather than independently monitor and adjust treatment plans. |
Explain procedures and discuss test results or prescribed treatments with patients.
15CI 5–25 · exposure 17 · augmentation 63 · importance 4.8/5 · click for rater detail
Explain procedures and discuss test results or prescribed treatments with patients.
15| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | OB/GYN is a high-touch specialty with significant liability exposure and strong professional norms around direct patient communication; adoption of autonomous AI for this task remains negligible in clinical practice. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare overall adopts AI slower than other professional services due to regulation, liability, and patient trust concerns, with pilots for patient communication tools more common than full production deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by drafting patient-friendly summaries, organizing test data, or suggesting explanatory frameworks that a physician then refines and delivers verbally, moderately improving preparation and documentation efficiency. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully help physicians prepare patient-friendly explanations, summarize test results, and draft follow-up materials, significantly aiding the communication task while the physician remains the primary communicator. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires direct patient interaction, empathetic communication, and responsiveness to individual concerns—capabilities that current AI systems cannot reliably replicate in a clinical setting. AI cannot autonomously conduct the nuanced, real-time dialogue needed to explain complex medical procedures or address patient anxieties. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can help draft explanations of procedures and test results, but delivering this in-person with empathy, answering follow-up questions, and adapting to patient emotional state requires human judgment and physical presence that current systems cannot fully replace. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and professional barriers exist: physicians have a duty to personally inform patients about procedures and treatments, informed consent requires direct physician-patient dialogue, and malpractice liability is high if a patient's understanding is inadequate or mediated by unsupervised AI. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Informed consent and treatment discussions in obstetrics/gynecology carry high liability and typically require a licensed physician to legally discuss and authorize treatment plans, especially given sensitive reproductive health contexts. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The oversight, customization, and validation burden to deploy AI for patient communication would exceed the cost savings from a physician's time, especially given the high liability exposure of miscommunication in obstetrics and gynecology. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While generating explanatory text is cheap, the actual patient-facing conversation still requires physician time for liability, nuance, and trust-building, keeping overall cost comparable to human-delivered care. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can generate templated explanations or summarize test results, no deployed product reliably handles the full task of explaining procedures and discussing treatment with patients in a clinical context. Narrow prototypes exist but lack the adaptability, liability protection, and integration into real OB/GYN workflows that production systems require. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Patient education chatbots and portals exist to explain test results, but no deployed product independently conducts the full clinical conversation about procedures and treatment options at scale in production OB/GYN settings. |
Conduct research to develop or test medications, treatments, or procedures to prevent or control disease or injury.
14CI 7–20 · exposure 17 · augmentation 63 · importance 3.7/5 · click for rater detail
Conduct research to develop or test medications, treatments, or procedures to prevent or control disease or injury.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While pharmaceutical and academic research organizations use AI for data analysis and drug discovery screening, the actual conduct of clinical research and treatment development remains heavily human-dependent. Adoption of AI in this domain is limited to specific phases (literature review, statistical analysis) rather than the full research pipeline. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Clinical and biomedical research adoption of AI is growing but remains at the pilot stage for hypothesis generation and data analysis; wide-scale autonomous research adoption in OB/GYN specifically is rare. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist researchers by accelerating literature synthesis, conducting statistical analyses, identifying patterns in trial data, and automating regulatory documentation—improving productivity on discrete research components. However, the core intellectual work of hypothesis generation, experimental design, and clinical interpretation remains fundamentally human. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI significantly aids literature review, data analysis, trial design optimization, and pattern detection in research data, meaningfully boosting researcher productivity while humans retain control over study design and interpretation. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Conducting research to develop or test medications, treatments, or procedures requires hypothesis generation, experimental design, clinical judgment, and interpretation of complex biological data—tasks that current AI cannot perform end-to-end without substantial human oversight and direction. AI can assist with literature review and data analysis, but cannot independently design, execute, or interpret clinical trials. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can assist with literature review, hypothesis generation, and data analysis, but designing, conducting, and validating original medical research including clinical trials requires human scientific judgment, ethical oversight, and physical/clinical execution that current AI cannot replace end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Clinical research is heavily regulated by IRBs, FDA, and other agencies that require human researcher accountability, informed consent processes, and licensed clinician oversight. Liability for adverse events and efficacy claims rests on human researchers and institutions, creating legal and regulatory barriers that prevent full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Medical research involving human subjects is heavily regulated (IRB approval, FDA oversight, licensed physician involvement), creating hard legal barriers that prevent full AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Research development and clinical testing are capital-intensive, require specialized human expertise (physicians, biostatisticians, regulatory specialists), and demand accountability that current AI systems cannot provide. The cost of AI-assisted analysis is negligible compared to the human researcher salaries, infrastructure, and regulatory compliance required. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply assist with data analysis and literature synthesis, but the overall research process still requires costly clinical infrastructure, trial staff, and regulatory compliance, keeping the all-in cost comparable to or higher than pure AI substitution claims suggest. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI tools exist for literature mining and statistical analysis of trial data, no deployed product reliably performs the full research-to-testing pipeline independently. Clinical research still depends on human researchers for protocol design, patient recruitment, adverse event assessment, and regulatory submission—making current AI applications narrow and supportive rather than autonomous. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tools (e.g., literature mining, drug candidate screening, statistical analysis platforms) are used in research support roles, but no deployed product independently conducts or validates OB/GYN clinical research at production scale. |
Plan, implement, or administer health programs in hospitals, businesses, or communities for prevention and treatment of injuries or illnesses.
14CI 3–25 · exposure 13 · augmentation 63 · importance 3.5/5 · click for rater detail
Plan, implement, or administer health programs in hospitals, businesses, or communities for prevention and treatment of injuries or illnesses.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare organizations are slow to adopt autonomous AI for high-stakes administrative and clinical planning functions due to regulatory constraints, liability concerns, and institutional conservatism. Pilot projects exist, but production-level autonomous health program administration remains rare. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administration adopts AI slowly relative to information/finance sectors due to regulatory complexity, liability concerns, and organizational inertia. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist physicians and administrators by providing data analysis, benchmarking against other programs, evidence synthesis, and draft document generation, enabling faster research and planning phases while humans retain decision authority and accountability. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully assist with epidemiological data analysis, drafting program proposals, literature review, and tracking outcomes, enhancing physician productivity in planning tasks. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Planning, implementing, and administering health programs requires strategic decision-making, stakeholder engagement, regulatory navigation, and clinical judgment that integrates complex organizational and medical knowledge. Current AI cannot autonomously perform the full scope of program administration, policy development, and clinical oversight required in this task. |
| Task automatability | claude-sonnet-5 | 2/5 | Program planning involves stakeholder coordination, resource allocation, policy judgment, and organizational politics that current AI cannot execute end-to-end, though it can assist with drafting and data analysis components.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Health program planning and administration in hospitals and clinical settings are heavily regulated by state and federal law, accreditation bodies, and institutional governance. Physicians must legally sign off on clinical aspects, and organizational liability for program failures remains with licensed professionals, creating hard barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Health program administration typically requires licensed physician leadership, institutional accountability, and regulatory compliance, creating strong barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI systems combined with mandatory expert human oversight, validation, and final authority for health program design and implementation remains comparable to or exceeds the cost of direct physician and administrator labor for these tasks. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply generate drafts or analyze data, but the overall task requires physician oversight, stakeholder buy-in, and implementation management that keep human labor costs dominant. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with literature review, data analysis, and draft templates for health programs, no deployed product reliably executes end-to-end health program planning and administration at scale. Existing products are narrow, lack the contextual and clinical decision-making capability, and require substantial human oversight and direction. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously plans or administers hospital/community health programs; this remains a human administrative and clinical leadership function with AI only as a supporting tool. |
Consult with or provide consulting services to other physicians.
13CI 5–20 · exposure 17 · augmentation 63 · importance 4.7/5 · click for rater detail
Consult with or provide consulting services to other physicians.
13| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of AI for autonomous physician-to-physician consulting is effectively non-existent in production settings, as legal and professional norms require a human physician to be the consulting party. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialist medicine, is a slow-adopting sector for autonomous AI decision-making due to regulatory and liability constraints, though AI-assisted tools are gradually being piloted. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by summarizing case literature, flagging relevant guidelines, or generating structured case summaries that support a physician's consulting process, though the consultation itself remains physician-driven. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist consulting physicians by summarizing literature, patient records, imaging, or suggesting differentials, improving speed and thoroughness while the physician retains responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Consulting with other physicians requires nuanced clinical judgment, contextual understanding of complex cases, and relationship-building that current AI systems cannot replicate end-to-end. AI cannot reliably substitute for the specialized expertise and accountability expected in peer consultation. |
| Task automatability | claude-sonnet-5 | 2/5 | Physician-to-physician consultation requires synthesizing complex patient-specific clinical judgment, liability, and nuanced communication that current AI cannot fully replicate end-to-end, though it can support parts like literature lookup or differential generation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and professional barriers exist: only licensed physicians can provide medical consulting, professional liability falls on the consulting physician, and medical boards require human accountability for clinical recommendations. Peer consultation is inherently a credentialed human activity. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Formal consultation between physicians is a licensed medical act carrying legal liability, credentialing, and malpractice implications, requiring a licensed physician to provide and sign off on the consult. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The loaded cost of a physician's consultation time remains significantly lower than the combined cost of AI systems, integration, verification, and liability oversight required to attempt autonomous consulting services. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools are cheap per query but cannot independently deliver a liable, board-certified consult, so the effective cost still requires a supervising OB/GYN, keeping overall cost comparable to human consultation. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in generating differential diagnoses or literature summaries for case discussion, no deployed product reliably performs peer consultation independently. Deployed systems lack the credibility, legal standing, and contextual depth required for genuine physician-to-physician consultation. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Clinical decision support and AI diagnostic tools exist but are not deployed as substitutes for physician-to-physician consults; they remain adjuncts used by the consulting physician rather than autonomous consultants. |
Refer patient to medical specialist or other practitioner when necessary.
6CI 0–11 · exposure 5 · augmentation 50 · importance 4.4/5 · click for rater detail
Refer patient to medical specialist or other practitioner when necessary.
6| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare remains a laggard sector for autonomous AI deployment due to regulatory burden, liability concerns, and the centrality of human clinical judgment to practice; referral decisions in particular are not being delegated to AI agents in production systems across the industry. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare overall adopts clinical-decision AI slowly due to regulatory, liability, and EHR integration constraints, with referral workflows seeing little automation in production. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by suggesting relevant specialists based on diagnosis, retrieving specialist availability and credentials, or flagging contraindications, thereby accelerating the practitioner's referral workflow—but the decision authority and patient communication remain with the physician. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help flag risk factors, summarize records, or suggest relevant specialists to consider, offering moderate assistance while the physician retains the referral decision. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Referral decisions require clinical judgment about patient complexity, specialist availability, and appropriateness of care—capabilities that demand medical knowledge and contextual reasoning current AI systems cannot reliably replicate end-to-end. The task is fundamentally a human clinical decision that cannot be automated to the 50% time-saving threshold without unacceptable liability risk. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a clinical judgment task requiring synthesis of patient history, exam findings, and specialist availability; AI cannot reliably perform the referral decision and coordination end-to-end today.}, |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Medical licensure laws require a licensed physician to make clinical referral decisions; liability, malpractice exposure, and regulatory standards (state medical boards, CMS, accreditation) all mandate that a qualified human clinician must authorize any patient handoff or specialist routing. This is a hard legal and professional barrier. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Referral decisions are a core licensed physician responsibility with direct liability and legal accountability, making this a hard-barrier task requiring physician sign-off. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task is low-volume per patient encounter and embedded in the physician's broader consultation workflow; automating it would require healthcare IT infrastructure investment and oversight mechanisms that exceed the savings of avoiding a single referral decision per visit. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Since AI cannot independently perform the task, any AI use still requires the physician's full involvement, so there is minimal cost savings versus the human doing it. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs this task autonomously; while EHR systems can flag referral prompts or route based on coded diagnoses, the clinical determination of when and to whom to refer remains a human practitioner function in all production healthcare settings. AI can assist with information retrieval but not make the referral decision itself. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously makes referral decisions in obstetric/gynecologic practice; at most decision-support tools flag possible specialist needs for physician review.ate |
Treat diseases of female organs.
5CI 3–7 · exposure 5 · augmentation 50 · importance 4.9/5 · click for rater detail
Treat diseases of female organs.
5| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While healthcare is digitizing, adoption of AI for autonomous treatment remains minimal; AI in OB/GYN is largely limited to diagnostic support tools under strict physician control. Regulatory and liability constraints keep velocity low for autonomous or near-autonomous clinical decision-making. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare overall adopts AI slowly for actual treatment delivery due to regulation and liability, though diagnostic aids are increasingly piloted. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI offers moderate assistance through imaging analysis, differential diagnosis support, and evidence retrieval that can inform a gynecologist's decision-making. However, augmentation is limited by the need for human clinical judgment, patient interaction, and procedural execution. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with diagnostics, imaging interpretation, treatment planning support, and documentation, improving efficiency while the physician remains the treating agent. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Treating diseases of female organs is a complex clinical task requiring hands-on medical intervention, diagnosis, judgment calls, and often surgical procedures. Current AI cannot perform diagnostic assessments, prescribe treatments, or execute any procedural interventions end-to-end in this domain. |
| Task automatability | claude-sonnet-5 | 1/5 | Treating diseases of female organs requires physical examination, surgery, prescribing, and hands-on procedures that current AI cannot perform end-to-end; this is not a text/data task amenable to full automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Treating diseases of female organs requires a licensed MD/DO and is governed by extensive regulatory requirements, medical malpractice liability, patient-contact mandates, and informed consent. Legal and ethical frameworks explicitly require a credentialed human physician to perform diagnosis and treatment. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Treating disease requires a licensed physician, involves surgery/prescriptions, and carries high liability, all legally requiring human authorization and performance. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task fundamentally requires a licensed physician whose multi-year training, malpractice insurance, and clinical decision-making authority represent high sunk and ongoing costs. AI inference and decision support cannot replace this human infrastructure. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing the actual treatment, so cost comparison favors the human physician entirely; AI tools add cost as adjuncts rather than replacing labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with imaging interpretation (ultrasound, MRI) and some diagnostic support, no deployed AI system can independently treat diseases or manage complex gynecological conditions. Clinical deployment remains limited to narrow support roles with mandatory physician oversight. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously treats gynecological disease; AI exists only as diagnostic support (e.g., imaging analysis) rather than a treatment-delivering system. |
Prescribe or administer therapy, medication, and other specialized medical care to treat or prevent illness, disease, or injury.
5CI 3–7 · exposure 5 · augmentation 75 · importance 4.7/5 · click for rater detail
Prescribe or administer therapy, medication, and other specialized medical care to treat or prevent illness, disease, or injury.
5| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare remains moderately digitized with fragmented EHR systems; adoption of AI in clinical decision-making is mostly in pilot or advisory roles. Regulatory caution, liability concerns, and physician skepticism slow deployment of autonomous systems in treatment decisions. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially high-liability specialties like OB/GYN, adopts AI cautiously and mainly for documentation or decision support, not for autonomous treatment decisions. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly augments physician productivity in this space through drug interaction alerts, evidence-based guideline recommendations, differential diagnosis suggestions, and real-time laboratory result interpretation, while the physician retains full decision authority and responsibility. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully assist by surfacing drug interactions, treatment guidelines, and risk flags, improving physician decision speed and accuracy while the physician remains fully in control. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Prescribing and administering therapy, medication, and specialized medical care requires complex clinical judgment, patient history integration, real-time assessment of patient response, and legal authority that only licensed physicians can exercise. Current AI cannot autonomously make these decisions at 50% time savings with equal safety. |
| Task automatability | claude-sonnet-5 | 1/5 | Prescribing and administering treatment requires direct physical examination, real-time clinical judgment, and legal accountability that current AI cannot perform end-to-end; AI can support diagnosis but not execute the full task. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Prescribing and administering medical treatment is legally restricted to licensed physicians (or nurse practitioners/PAs under supervision). Malpractice liability, regulatory oversight by medical boards, FDA authority over medications, and patient safety requirements create hard legal and liability barriers to autonomous AI automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Prescribing medication and administering medical care legally requires a licensed physician; strict regulatory, licensing, and liability frameworks make this a hard-barrier task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI decision-support tools require significant infrastructure, EHR integration, liability management, and human oversight. The fully-loaded cost of deploying, maintaining, and legally standing behind AI recommendations remains higher than the marginal time saved versus a physician's judgment. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot independently perform prescribing/administration, the physician cost remains fully required, making AI substitution cost irrelevant to replacing this task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in drug interaction checking and guideline-based recommendations, no deployed system reliably performs end-to-end prescribing or treatment administration autonomously. AI tools exist as decision-support aids in some EHRs, but the clinician remains the legal decision-maker and operator. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously prescribes or administers medication/therapy in obstetrics/gynecology; clinical decision-support tools exist but require physician sign-off and action. |
Care for and treat women during prenatal, natal, and postnatal periods.
1CI 0–3 · exposure 0 · augmentation 63 · importance 4.9/5 · click for rater detail
Care for and treat women during prenatal, natal, and postnatal periods.
1| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Despite healthcare digitization, obstetric delivery and prenatal care remain fundamentally human-dependent and have shown minimal automation adoption. Regulatory, safety, and liability concerns mean even AI-assisted pilots are rare in active labor management. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially obstetric care, is a highly regulated, physical-contact-intensive sector with slow AI adoption for core clinical delivery tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI tools can assist with fetal heart rate monitoring analysis, risk prediction models, and radiology interpretation to support clinical decision-making, but the human physician must remain central to all diagnostic and treatment decisions given the high stakes of pregnancy and childbirth. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools meaningfully assist with fetal monitoring analysis, risk stratification, ultrasound interpretation, and documentation, improving physician efficiency while humans remain central to care delivery. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task involves direct physical examination, clinical decision-making under uncertainty, and hands-on medical intervention during pregnancy and childbirth. Current AI systems cannot perform pelvic exams, manage labor complications in real time, or deliver babies—the core clinical functions of this role. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical examination, delivery of babies, surgical intervention (C-sections), and hands-on care across many months that cannot be performed end-to-end by current AI systems. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Obstetric and gynecologic care is heavily regulated and requires a licensed MD/DO. Malpractice liability is substantial, informed consent depends on direct patient relationships, and labor/delivery management legally requires a physician present. Hard regulatory and legal barriers prevent substitution. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Medical licensure, malpractice liability, and legal requirements for physician-attended childbirth create hard, high barriers to any substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Obstetric care is complex and high-stakes; the cost of deploying AI systems with sufficient safety oversight, regulatory compliance, and malpractice insurance would far exceed the loaded wage of a physician in most healthcare systems. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical delivery of care and physician liability, so there is no viable AI-only cost comparison; human labor remains essential. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product performs end-to-end obstetric and gynecologic care. While AI can assist with imaging interpretation or risk stratification, no system reliably manages the full scope of prenatal monitoring, labor management, and postpartum care independently. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product provides comprehensive prenatal, natal, and postnatal care; AI tools exist only as narrow decision-support or imaging aids alongside physicians. |
Perform cesarean sections or other surgical procedures as needed to preserve patients' health and deliver babies safely.
0CI 0–0 · exposure 0 · augmentation 38 · importance 4.8/5 · click for rater detail
Perform cesarean sections or other surgical procedures as needed to preserve patients' health and deliver babies safely.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Obstetric surgery remains a hands-on, human-performed clinical task with minimal autonomous automation adoption in production; sectors remain conservative due to safety and liability concerns. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Surgical practice is highly resistant to AI-driven task displacement; adoption in this domain is confined to assistive tools, not autonomous procedure execution, and progresses very slowly due to safety and regulatory constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Surgical robots and imaging AI can assist surgeons by stabilizing instruments or enhancing visualization, but these are narrow aids rather than transformative productivity multipliers for the core task of performing the procedure itself. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI and robotic-assisted surgical systems (e.g., image guidance, robotic arms controlled by surgeons) can enhance precision and decision support during procedures, though the human surgeon remains fully in control. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Cesarean sections and obstetric surgery require real-time manual dexterity, intraoperative decision-making based on visual and tactile feedback, and immediate response to complications. Current AI systems cannot perform surgical procedures end-to-end; surgical robots require direct human control and cannot operate autonomously. |
| Task automatability | claude-sonnet-5 | 1/5 | Physical surgery requiring manual dexterity, real-time judgment, and hands-on intervention cannot be performed end-to-end by current AI systems; no off-the-shelf system executes surgical procedures autonomously.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Surgical procedures require a licensed physician to perform or directly oversee them; malpractice liability, patient safety regulations, and legal requirements create hard barriers to automation or delegation away from credentialed medical professionals. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Performing surgery legally requires a licensed physician; liability, malpractice law, and patient safety regulations make this one of the most protected tasks against automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The infrastructure, maintenance, and oversight costs of any surgical automation system, combined with necessary human supervision and liability, would far exceed the cost of a trained obstetrician-gynecologist performing the surgery. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this task, so cost comparison favors the human surgeon by default; any surgical robotics involve high capital and human oversight costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product performs cesarean sections or obstetric surgery autonomously. Surgical robotics assist human surgeons but do not execute procedures independently or meet the threshold of reliable, unassisted task performance. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs cesarean sections or comparable surgeries autonomously; surgical robots today are human-controlled tools, not autonomous performers of the procedure. |
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.