Tutors
25-3041.00Median wage $43,350/yr175,070 employed (US)Rank #94 of 923 scored · top 10% by substitution
Instruct individual students or small groups of students in academic subjects to support formal class instruction or to prepare students for standardized or admissions tests.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution47
Exposure42
Augmentation74
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
19 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
21%
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%42
panel mean rating 2.7/5 → substitution pressure 42/100
Technical feasibility todayw 20%42
panel mean rating 2.7/5 → substitution pressure 42/100
Cost vs. human wagew 15%57
panel mean rating 3.3/5 → substitution pressure 57/100
Adoption barriersw 20%inverted — strong barriers lower the score58
panel mean rating 2.7/5 (barrier strength) → substitution pressure 58/100
Sector adoption velocityw 10%41
panel mean rating 2.7/5 → substitution pressure 41/100
Task breakdown (19 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.
Develop teaching or training materials, such as handouts, study materials, or quizzes.
79CI 76–81 · exposure 75 · augmentation 100 · importance 3.4/5 · click for rater detail
Develop teaching or training materials, such as handouts, study materials, or quizzes.
79| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | EdTech and online learning sectors show moderate adoption of AI-generated materials (pilots and early production use); however, traditional K–12 and higher education institutions lag, with many tutors still using manual or template-based workflows. Uptake is growing but not yet dominant in most tutoring contexts. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Education technology adoption is moderate—many tutors and teachers experiment with AI content generation tools, but institutional and individual uptake varies widely and is not yet universal. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI dramatically amplifies tutor productivity by rapidly generating drafts of handouts, quizzes, and study guides that tutors refine and customize; this keeps human judgment in the loop while reducing content-creation time by 70–90%. This is one of the clearest augmentation wins in education. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI substantially speeds up drafting quizzes, handouts, and study guides, letting tutors focus on customization, delivery, and pedagogical judgment while AI produces first drafts. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can generate high-quality handouts, study materials, and quizzes at scale with minimal human input; LLMs excel at creating varied question formats and explanatory content. However, customization to specific curricula, student levels, and learning outcomes typically requires human review and iteration, preventing full end-to-end automation. |
| Task automatability | claude-sonnet-5 | 4/5 | LLMs can generate handouts, study guides, practice quizzes, and explanations quickly given a topic and level, requiring mostly review/editing rather than creation from scratch. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or legal barriers exist to automating material creation; most institutions do not require licensed educator sign-off on the generation process itself, though human review is common practice. Adoption is primarily limited by inertia and quality concerns rather than hard restrictions. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing or legal requirement mandates a human create teaching materials; there's no liability barrier for draft content that a tutor reviews before use. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI inference cost for generating materials is negligible (pennies per task), while a tutor creating equivalent handouts and quizzes demands hours of skilled labor at $20–60+ per hour loaded cost. The cost advantage is substantial—typically 50–100× cheaper. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Generating draft materials via AI costs cents to dollars versus the hourly wage of a tutor spending significant time on manual content creation. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed products (ChatGPT, Claude, specialized EdTech platforms) reliably generate tutoring materials in production; quality is generally sufficient for classroom use. Minor limitations exist in ensuring pedagogical alignment and avoiding generic content, but systems demonstrably perform this task at scale. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Products like ChatGPT, Khanmigo, and quiz-generation tools are already used by educators and tutors in production to draft materials, though accuracy and alignment to specific curricula still need human review. |
Research or recommend textbooks, software, equipment, or other learning materials to complement tutoring.
79CI 76–81 · exposure 75 · augmentation 100 · importance 2.8/5 · click for rater detail
Research or recommend textbooks, software, equipment, or other learning materials to complement tutoring.
79| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Educational technology adoption is steady but fragmented; independent tutors and small tutoring businesses have lower adoption rates than larger platforms or institutional tutoring programs, resulting in moderate velocity overall. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Education is adopting AI tools steadily but unevenly; many tutors and tutoring platforms use AI assistance already, though widespread deep integration is still developing. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI-powered research and recommendation tools significantly enhance tutor productivity by rapidly surfacing curated options, enabling tutors to spend more time customizing materials and working directly with students rather than hunting for resources. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI substantially speeds up and broadens the search for suitable materials, letting tutors quickly get options and rationale while retaining final selection judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can systematically research and recommend learning materials by analyzing curriculum standards, student needs, and resource databases; this could save 60–80% of time on research and initial recommendation generation, though human judgment on fit for specific students remains important. |
| Task automatability | claude-sonnet-5 | 4/5 | AI can research and recommend learning materials effectively by synthesizing curriculum needs, reviewing reviews/specs, and generating tailored lists, saving substantial time versus manual research.dependant judgment on subject fit still helps but most legwork is automatable. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Minimal regulatory or licensing barriers exist for recommending materials; tutors typically retain discretion on selections, and organizational friction is low in most tutoring settings. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing or liability barrier prevents recommending learning materials; it's a low-stakes advisory task with no regulatory requirement for human sign-off. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI inference and integrated recommendation tools cost negligibly compared to a human tutor spending 1–2 hours per week researching and vetting materials, representing well over 10× cost advantage. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Generating recommendations via an LLM costs fractions of a cent compared to the tutor's billable research time, making AI drastically cheaper for this subtask. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed systems (LLMs, educational recommendation engines, search-augmented tools) already perform materials research and recommendation at scale in educational platforms; products exist in production though recommendations often require tutor review for context-specificity. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Products like ChatGPT, Khanmigo, and other ed-tech recommendation tools already provide curated material suggestions in production, though quality varies by subject specificity. |
Schedule tutoring appointments with students or their parents.
76CI 67–84 · exposure 70 · augmentation 75 · importance 4.0/5 · click for rater detail
Schedule tutoring appointments with students or their parents.
76| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Educational technology platforms increasingly incorporate automated scheduling (Calendly, Acuity, learning management systems), but uptake among independent tutors and smaller tutoring centers remains mixed. Adoption is rising but not yet sector-wide standard practice. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Scheduling automation is broadly adopted across education, tutoring services, and professional services generally, with tools like Calendly and Acuity in widespread production use.4 |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI scheduling assistants materially improve tutor productivity by handling availability synchronization, reminders, and routine confirmations, freeing the tutor for instruction prep. The human tutor remains in control of final approval and parent communication tone. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Even where humans remain involved, AI scheduling assistants substantially reduce back-and-forth communication and administrative burden for tutors and parents.4 |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can largely automate scheduling logic via calendar integration, availability matching, and email/SMS confirmation, achieving substantial time savings. However, handling complex constraints (last-minute changes, parent preferences, multiple student coordination) may still require some human judgment, falling short of full end-to-end automation at equal quality. |
| Task automatability | claude-sonnet-5 | 4/5 | Scheduling is a well-structured coordination task easily handled by AI scheduling assistants and chatbots that check calendars, propose times, and confirm appointments with minimal human input.4 |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Scheduling automation has few legal or licensing barriers; no regulation requires a human to book appointments. The main friction is organizational adoption inertia and preference for human touch in parent communication, which are moderate but not hard constraints. |
| Adoption barriers | claude-sonnet-5 | 1/5 | No licensing, liability, or regulatory requirement mandates human involvement in scheduling; it's a purely administrative task with negligible friction.1 |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Scheduling automation (calendar APIs, chatbots, email systems) has minimal per-task marginal cost compared to a tutor's loaded hourly wage for handling administrative scheduling. Even with oversight overhead, the cost ratio is heavily in AI's favor. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated scheduling tools cost a few dollars per month versus paying a human tutor or administrator's time to manually coordinate appointments.5 |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Calendar scheduling automation and appointment-booking systems exist and are deployed, but most tutoring services still rely on hybrid workflows with human verification. Production reliability is moderate; edge cases and parent communication preferences often require manual intervention. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Mature scheduling products (Calendly, AI chat-based schedulers, CRM automation) are deployed at scale across tutoring companies and educational platforms today.4 |
Prepare lesson plans or learning modules for tutoring sessions according to students' needs and goals.
76CI 59–92 · exposure 70 · augmentation 100 · importance 3.6/5 · click for rater detail
Prepare lesson plans or learning modules for tutoring sessions according to students' needs and goals.
76| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | EdTech and tutoring platforms are among the fastest adopters of generative AI. ChatGPT, Claude, and specialized tutoring platforms report heavy use for lesson planning; many independent tutors and small tutoring firms have integrated AI planning tools into daily workflows. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Education technology adoption is growing steadily with many tutoring platforms integrating AI planning tools, but many individual tutors still rely on manual or templated methods. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | Even when a human tutor remains the primary lesson designer, AI dramatically accelerates iteration, generates differentiated scaffolding, and surfaces activity ideas, allowing tutors to focus on pedagogical judgment and student engagement rather than administrative drafting work. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is widely used today to rapidly generate drafts, practice problems, and structured modules that tutors then customize, substantially speeding up prep work while the tutor retains final control. |
| Task automatability | claude-haiku-4-5-20251001 | 5/5 | AI can generate comprehensive lesson plans and learning modules tailored to student profiles, learning objectives, and proficiency levels end-to-end with >50% time savings. Large language models and educational AI platforms already deliver full-draft lesson structures, activities, assessments, and resource recommendations, dramatically reducing the manual planning burden. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft lesson plans and adapt content quickly, but tailoring to a specific student's evolving needs, learning style, and goals still requires human judgment and iteration to reach equal quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few legal or regulatory barriers prevent AI lesson-plan generation; no license or mandatory human sign-off is required. Light barriers include customer expectation of human personalization and organizational inertia, but these are soft friction, not hard legal constraints. |
| Adoption barriers | claude-sonnet-5 | 1/5 | There is no licensing or regulatory requirement mandating that a human create lesson plans, and tutoring is a largely unregulated profession. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI-generated lesson planning costs pennies per session via API calls or subscription, compared to $15–50+ hourly loaded cost for a human tutor to prepare equivalent materials. The cost gap is at least 10–50x in favor of AI. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Generating a lesson plan draft via AI costs a fraction of a cent to a few cents versus the tutor's hourly wage for prep time, though oversight time reduces the net savings somewhat. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Multiple deployed products (ChatGPT, Claude, specialized EdTech platforms) routinely generate lesson plans and learning modules in production. While human review and customization remain standard practice, the core task of drafting pedagogically sound, differentiated modules is now reliably automated at scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Products like Khan Academy's Khanmigo and various AI lesson planners are deployed and used by tutors, but they still require significant human review and customization for accuracy and fit. |
Review class material with students by discussing text, working solutions to problems, or reviewing worksheets or other assignments.
69CI 59–79 · exposure 62 · augmentation 88 · importance 4.7/5 · click for rater detail
Review class material with students by discussing text, working solutions to problems, or reviewing worksheets or other assignments.
69| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | EdTech adoption of AI tutoring is already rapid, with widespread deployment in online tutoring platforms, learning management systems, and consumer apps. K–12 and higher education institutions are piloting and scaling AI-assisted review tools at accelerating pace. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | EdTech and AI tutoring products (Khanmigo, Duolingo, etc.) are seeing growing adoption, but human tutoring remains dominant in most settings with pilots more common than full deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI powerfully augments human tutors by instantly generating worked solutions, creating adaptive review materials, and enabling 24/7 availability, while the human tutor retains oversight, motivation, and personalized guidance—a canonical human-in-the-loop augmentation scenario. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can generate practice problems, explain solutions, and provide instant feedback, substantially aiding a tutor's ability to review material efficiently. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can autonomously review class material, discuss concepts, work through problem solutions, and provide feedback on assignments with high quality and efficiency, meeting the 50% time-saving threshold. However, the interactive, personalized engagement component and need for human judgment on individual learning gaps prevent a full 5-rating. |
| Task automatability | claude-sonnet-5 | 3/5 | AI tutoring tools (e.g., chatbots) can explain text and walk through problem solutions with reasonable quality, but replicating the full interactive, adaptive rapport of a human tutor across a session still requires significant human oversight for many students.4 |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few hard legal barriers exist; tutoring is not a licensed profession in most jurisdictions. Main friction is parental/institutional preference for human interaction, curriculum alignment, and quality-assurance concerns, but these are not legal or regulatory blockers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing is required to tutor in most contexts, though some parents/students prefer human interaction and schools may require certified staff for formal instruction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI tutoring inference and integration costs are negligible per session compared to the loaded wage of a human tutor (typically $20–50/hour), making AI at least an order of magnitude cheaper for equivalent task delivery at scale. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI tutoring subscriptions cost a small fraction of hourly human tutoring rates for comparable content review. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed AI tutoring systems (e.g., ChatGPT, specialized ed-tech platforms) reliably perform text discussion, problem-solving, and worksheet review in production. Error rates on straightforward material are low, though performance varies with subject complexity and student-AI interaction quality. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Wait, fix format |
Administer, proctor, or score academic or diagnostic assessments.
64CI 54–74 · exposure 62 · augmentation 63 · importance 2.9/5 · click for rater detail
Administer, proctor, or score academic or diagnostic assessments.
64| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | K–12 and higher-ed institutions are rapidly adopting automated scoring and proctoring tools; deployment is now standard in many online and hybrid learning environments. Tutoring platforms (Chegg, Tutor.com) increasingly integrate AI scoring. Adoption is driven by cost, scale, and digitization in education sectors. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | EdTech and online education have adopted AI-based proctoring and scoring tools significantly, but adoption is uneven across in-person tutoring markets and diagnostic assessment contexts, with many tutors still using manual methods. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI assists tutors by auto-scoring routine assessments and flagging outliers, freeing time for diagnostic feedback and personalized instruction. However, the augmentation is modest—scoring itself does not significantly raise the tutor's cognitive productivity once automated; benefit comes mainly from time liberation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools substantially help tutors by auto-scoring quizzes, flagging error patterns, and generating diagnostic reports, freeing tutors to focus on interpretation and remediation planning. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Scoring assessments, especially multiple-choice and standardized tests, can be nearly fully automated by current AI systems. Proctoring (identity verification, plagiarism detection, test-environment monitoring) is increasingly automatable. Administering assessments is partially automatable (scheduling, distribution, data collection). Only adaptive, real-time test administration with complex judgment faces obstacles. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can score standardized or short-answer/essay assessments and administer computer-based tests reasonably well, but proctoring (monitoring for integrity) and administering in-person diagnostic assessments still require substantial human involvement, especially for younger students or special needs diagnostics. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Regulatory and institutional barriers exist in high-stakes (standardized, college entrance, professional certification) assessment contexts where human proctoring or independent verification may be legally or professionally mandated. Lower-stakes classroom assessments face minimal legal barriers but encounter institutional friction and educator preference for human oversight. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Some barriers exist around academic integrity standards, institutional policy requiring human oversight, and diagnostic assessments often requiring credentialed judgment, but no strict licensing requirement universally applies to tutors performing this task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated scoring and proctoring are orders of magnitude cheaper than human tutors or test administrators at scale—a single AI instance handles thousands of assessments per month at near-zero marginal cost, versus labor at $20–40/hour per assessment. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated scoring and proctoring software is far cheaper per assessment than paying a human tutor to manually score or supervise, especially at scale for standardized content. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature products exist for automated scoring (Turnitin, Gradescope, LMS systems) and basic proctoring (Honorlock, Proctorio in production). AI scoring of short-answer and open-ended responses has demonstrated reliability on benchmark sets, though still with error rates requiring human review in high-stakes contexts. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Products like AI proctoring software (e.g., ProctorU, Examity) and automated essay/short-answer scoring tools are deployed at scale, but they have known error rates, bias issues, and limited applicability to diagnostic/clinical-style assessments common in tutoring. |
Maintain records of students' assessment results, progress, feedback, or school performance, ensuring confidentiality of all records.
62CI 49–76 · exposure 62 · augmentation 75 · importance 3.8/5 · click for rater detail
Maintain records of students' assessment results, progress, feedback, or school performance, ensuring confidentiality of all records.
62| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 5/5 | Record-keeping automation is nearly universal in formal education; LMS and SIS platforms have achieved deep penetration across K–12 and higher education sectors, with widespread digital record maintenance already standard practice. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Education technology adoption is moderate—many tutoring platforms use digital recordkeeping tools, but full AI-driven automation of confidential records remains uneven across the sector. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-powered dashboards and analytics systems substantially assist tutors by generating automated progress summaries, flagging performance trends, and surfacing actionable insights from student data, enabling educators to spend less time on manual entry and more on intervention. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can significantly streamline record organization, feedback drafting, and progress tracking, greatly aiding tutors while they retain responsibility for confidentiality and accuracy. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can reliably capture, organize, and maintain structured records of assessment results and performance metrics with near-complete automation. The confidentiality requirement is handled through technical access controls and encryption rather than task complexity, achieving well over 50% time savings compared to manual record-keeping and data entry. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft, organize, and summarize progress notes and assessment data, but confidentiality management and final recordkeeping oversight still require human control, limiting full end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Schools and tutoring organizations have established compliance requirements around FERPA and data privacy that mandate oversight, audit trails, and human sign-off on certain record changes. These create moderate friction but do not legally require a human to perform the recordkeeping itself, only to authorize and audit it. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Confidentiality and data protection laws (e.g., FERPA) impose strict requirements on who can access and manage student records, creating substantial regulatory and liability barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | LMS and SIS platforms cost a small fraction of a tutor's hourly wage to maintain records for dozens of students, and bulk administrative solutions are orders of magnitude cheaper than manual data entry and filing by qualified personnel. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated record management systems are inexpensive relative to tutor time spent on manual documentation, though secure storage and compliance infrastructure add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed learning management systems (Canvas, Blackboard, Google Classroom) and student information systems routinely perform automated record maintenance, data logging, and performance tracking in thousands of schools and tutoring organizations. These products handle the core recordkeeping reliably at scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | EdTech platforms and LMS tools already generate and store student progress records with some AI-assisted summarization, but reliable, secure end-to-end automation of confidential recordkeeping is not yet universal. |
Teach students study skills, note-taking skills, and test-taking strategies.
59CI 41–76 · exposure 50 · augmentation 88 · importance 4.5/5 · click for rater detail
Teach students study skills, note-taking skills, and test-taking strategies.
59| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | EdTech adoption is growing steadily with pilots and some production use in schools and test-prep sectors, but most tutoring still remains human-delivered; uptake is middling rather than fast and deep compared to information-sector AI adoption. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | EdTech adoption is growing steadily with many pilots and increasing use of AI tutoring tools in schools and online tutoring platforms, though full replacement of tutors remains uncommon. |
| Augmentation potential | claude-haiku-4-5-20251001 | 5/5 | AI can powerfully augment human tutors by generating personalized study plans, diagnostic assessments, automated feedback on practice tests, and targeted skill-building exercises, freeing tutors to focus on mentorship and deeper conceptual questions while dramatically raising productivity. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can generate practice tests, structured note templates, and personalized study plans that meaningfully help tutors on this task, while the tutor still manages student engagement and adaptation. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | AI can fully automate structured instruction in study techniques, note-taking frameworks, and test-taking strategies through interactive tutoring systems, practice problem generation, and personalized feedback—easily meeting the 50% time-saving bar for the core content delivery and skill-building components of this task. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can generate content about study strategies but effective teaching requires diagnosing individual learning gaps, motivating students, and adapting delivery in real-time interaction, which current systems only partially replicate.4o and similar chatbots can produce generic advice but not the personalized coaching loop. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or licensure barriers specifically govern teaching study skills and test strategies; institutional adoption may face some organizational inertia and teacher resistance, but nothing prevents substitution or hybrid deployment in practice. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement for teaching study skills specifically, though schools may have policies on tutor qualifications; minimal regulatory barrier to AI-assisted delivery. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | AI tutoring systems cost a fraction of a human tutor's loaded hourly wage after initial development, serving unlimited students with negligible marginal cost per additional learner, achieving an order-of-magnitude cost advantage. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-based tutoring content generation is cheap per unit, but achieving comparable engagement and personalization to a human tutor still requires oversight and iteration, narrowing the cost advantage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature products (e.g., intelligent tutoring systems, adaptive learning platforms, and conversational AI tutors) reliably deliver study-skill and test-strategy instruction in production environments; reliability is high for well-defined pedagogical content, though real-time adaptation to individual student needs remains somewhat variable. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tutoring products (e.g., Khanmigo, Duolingo Max) offer some study-skill coaching but are narrow in scope and not yet reliable substitutes for personalized skill-building across diverse student needs. |
Provide private instruction to individual or small groups of students to improve academic performance, improve occupational skills, or prepare for academic or occupational tests.
52CI 45–59 · exposure 42 · augmentation 88 · importance 4.5/5 · click for rater detail
Provide private instruction to individual or small groups of students to improve academic performance, improve occupational skills, or prepare for academic or occupational tests.
52| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | AI tutoring is growing in edtech and test-prep sectors, but adoption remains concentrated in tech-forward segments and supplementary use. Traditional tutoring markets (local tutors, test prep centers) show slower AI integration; most tutoring remains human-delivered. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | EdTech adoption is growing steadily with many pilots and consumer products, but private tutoring remains a fragmented, small-business-dominated sector with uneven digitization compared to fast-moving professional service sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI excels at assisting human tutors by generating personalized problem sets, providing instant explanations, identifying knowledge gaps, and handling administrative tasks like scheduling and progress tracking, substantially multiplying what one tutor can manage. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI tools significantly enhance human tutors' capabilities by generating practice materials, explaining concepts in multiple ways, and providing instant feedback, allowing tutors to focus more on personalized guidance and motivation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate explanations, practice problems, and assess understanding, tutoring requires real-time adaptation to a student's emotional state, misconceptions, and learning pace—tasks that demand human judgment and rapport-building. Current systems lack the contextual awareness and interactive flexibility to reliably replace a human tutor end-to-end. |
| Task automatability | claude-sonnet-5 | 3/5 | AI tutoring systems can explain concepts, generate practice problems, and adapt to student responses for many subjects, but building rapport, reading nonverbal cues, and motivating disengaged students still require significant human involvement. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | No legal requirement for human licensure in most tutoring contexts, but significant parental and institutional preference for human contact, concerns over screen time, and liability questions around AI-only instruction create moderate friction to full substitution. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing is typically required for tutoring (unlike teaching credentials for classroom instruction), though parents/students often prefer human interaction and trust, creating moderate friction rather than hard legal barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Deployed AI tutoring per-session cost is dramatically lower than human tutors (pennies vs. tens of dollars per hour), though integration and content curation add overhead. The ratio favors AI substantially when amortized across many students. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI tutoring subscriptions cost a small fraction of hourly human tutoring rates, especially for standardized test prep and homework help, though some oversight and platform costs remain. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | AI tutoring products exist and are deployed (e.g., ChatGPT tutoring, Khan Academy's AI features), but they show material limitations in handling complex subjects, detecting frustration, and providing genuinely personalized scaffolding at scale. Production systems work for narrow domains (math drill, test prep) but struggle with open-ended instruction. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Deployed AI tutoring products (e.g., Khanmigo, various adaptive learning platforms) exist and are used in classrooms and homes, but they still have accuracy issues in complex subjects and limited ability to handle emotional/motivational aspects of tutoring. |
Participate in training and development sessions to improve tutoring practices or learn new tutoring techniques.
49CI 30–67 · exposure 41 · augmentation 75 · importance 4.2/5 · click for rater detail
Participate in training and development sessions to improve tutoring practices or learn new tutoring techniques.
49| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | EdTech and tutoring platforms are adopting AI-driven coaching and training tools at moderate pace, with growing pilot programs for AI-assisted staff development; however, uptake remains uneven across small tutoring centers and larger chains, and many tutors still prefer traditional PD workshops. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Education sector adoption of AI for professional development is emerging but slow and uneven, mostly pilot programs for content generation rather than full training replacement. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly enhances tutoring practice by providing on-demand scenario feedback, personalized technique suggestions, and unlimited practice simulations that tutors can learn from asynchronously; this transforms individual development while the tutor remains active in reflection and skill refinement. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully augment this task by generating personalized training materials, simulating student scenarios, and offering feedback on tutoring techniques, enhancing the tutor's learning experience. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | A significant portion of this task—curating learning materials, generating training content, simulating tutoring scenarios, and providing structured feedback on tutoring techniques—can be automated with current AI systems, achieving substantial time savings. However, the live collaborative and reflective components of genuine professional development sessions require some human interaction, preventing a full end-to-end 5 rating. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can supply training content and simulate scenarios but attending, engaging with, and applying live professional development sessions requires human participation and social learning that current AI cannot substitute end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few regulatory or legal barriers prevent AI-assisted training, and tutoring organizations have strong economic incentives to reduce professional development costs; organizational inertia around in-person cohort learning and individual tutor preferences for human mentorship represent the main friction, but these are soft barriers. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing barrier prevents AI-assisted training content, but the task inherently requires human presence and engagement in professional development, limiting substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-generated training materials, simulations, and feedback mechanisms cost a fraction of human-led professional development sessions; at scale, the per-session cost of AI-driven training is substantially lower than hiring trainers or facilitators, though some human oversight still requires payment. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Since AI cannot substitute for the human's participation itself, cost comparisons mostly apply to content creation tools, which are cheap but don't replace the core task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | AI products can generate training content, create practice scenarios, and deliver instructional materials reliably, but deployed systems for comprehensive tutoring skill assessment and adaptive coaching exist in limited form and often require human oversight to ensure contextual appropriateness and quality feedback. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | There are no deployed products that autonomously 'participate' in training as a tutor's stand-in; AI is only used to create supplementary training materials or coaching aids. |
Communicate students' progress to students, parents, or teachers in written progress reports, in person, by phone, or by email.
41CI 30–52 · exposure 42 · augmentation 75 · importance 3.9/5 · click for rater detail
Communicate students' progress to students, parents, or teachers in written progress reports, in person, by phone, or by email.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Most tutoring settings—private tutors, small tutoring centers, and traditional schools—have not widely deployed AI for progress reporting. Adoption remains limited and pilot-focused rather than embedded in production workflows at scale. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Education, especially individual tutoring, is a low-digitization sector with slow, uneven AI adoption for parent/student communication compared to finance or professional services. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-assisted drafting, template generation, and data summarization can meaningfully accelerate the composition of progress reports and emails, allowing tutors to focus on personalization and relationship-building. This is a strong augmentation scenario where humans remain in control. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially speeds up drafting progress reports and emails, letting tutors personalize and send communications faster while retaining control over tone and content. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can draft progress reports from structured data (grades, test scores, attendance), the task requires nuanced assessment of individual student development and personalized communication that meaningfully exceeds 50% time savings at equal quality. A human tutor must still review, contextualize, and customize the message for each recipient. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft written progress reports and email summaries competently from performance data, but in-person and phone communication requiring rapport and real-time judgment resist full automation.: |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Tutors and schools have professional and legal accountability for the accuracy and appropriateness of progress reports, and parents expect human judgment and accountability in communication about their child's development. Liability and trust concerns create strong friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement blocks AI-assisted communication, though parents and schools often expect personal human contact for sensitive student feedback, creating moderate preference-based friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI can reduce composition time for written reports, but the integration, customization per student, and required human review mean cost remains comparable to or only moderately lower than direct human composition, especially for in-person and phone communication. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Drafting written reports via AI is cheap, but phone/in-person communication still requires a human tutor's time, keeping blended cost roughly comparable rather than dramatically cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | AI tools can generate templated progress summaries and assist with written reports (e.g., LLM-based drafting), but no deployed system reliably replaces the full task end-to-end across varied student profiles and communication modes without substantial human oversight and revision. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | AI writing tools and ed-tech platforms already generate progress summaries and emails, but no mature product autonomously conducts phone or in-person progress conversations reliably at scale. |
Assess students' progress throughout tutoring sessions.
37CI 25–50 · exposure 30 · augmentation 75 · importance 4.7/5 · click for rater detail
Assess students' progress throughout tutoring sessions.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Ed-tech companies and K–12 districts increasingly deploy automated progress tracking and analytics, but adoption remains uneven; many tutoring services (especially independent tutors and specialized instruction) still rely primarily on human assessment; pilots are common, but deep AI-driven replacement of tutor judgment is not yet widespread. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Education/tutoring sector adoption of AI for continuous student assessment is still mostly pilot-stage, with slower uptake compared to sectors like finance or software. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-powered dashboards and real-time analytics can significantly assist tutors by surfacing patterns, flagging gaps, and recommending next steps, allowing tutors to focus their judgment on deeper interpretation and personalized intervention. This is a strong augmentation use case where AI handles data aggregation and human tutors direct strategy. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools (adaptive quizzes, analytics dashboards, performance tracking software) can meaningfully help tutors monitor student progress and identify weak areas, improving efficiency while tutor remains central to the judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can extract basic progress signals (e.g., correct/incorrect answers, response time) from structured assessments, but cannot reliably evaluate the nuanced, context-dependent progress that tutors assess—including conceptual understanding, effort, motivation, and readiness to advance. Significant human judgment remains essential. |
| Task automatability | claude-sonnet-5 | 2/5 | Assessing progress requires ongoing judgment about comprehension, motivation, and adaptive pacing that goes beyond what current AI can do end-to-end without human oversight in a live tutoring context.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: parents and students typically expect human judgment and feedback; liability concerns arise if AI-only assessments drive placement or intervention decisions; many tutoring contexts (especially specialized or high-stakes) have implicit or explicit expectations that a qualified human must evaluate progress. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement specifically for assessing progress, but tutoring often involves personalized human judgment and relationship-based trust that creates some friction to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | Automated progress tracking (via learning management systems and analytics) incurs minimal marginal cost per student once deployed, making it substantially cheaper than hiring tutors to manually assess; however, it does not fully replace the tutor's assessment role, so direct cost comparison is partial. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Building or licensing an assessment system with sufficient reliability and integration into a live tutoring session still requires meaningful human oversight cost, so savings versus a human tutor doing this informally are modest. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Products exist (learning analytics dashboards, adaptive tutoring systems) that track and visualize student performance metrics in real time, but these typically measure narrow outputs (quiz scores, problem completion) rather than holistic progress assessment as a tutor would conduct it. Deployment is widespread in ed-tech but with limited scope. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some AI tutoring products (e.g., adaptive learning platforms) can track quiz performance and flag gaps, but reliable holistic assessment of a student's real-time progress in a human-led session is not yet a mature deployed capability. |
Provide feedback to students, using positive reinforcement techniques to encourage, motivate, or build confidence in students.
37CI 29–45 · exposure 25 · augmentation 75 · importance 4.9/5 · click for rater detail
Provide feedback to students, using positive reinforcement techniques to encourage, motivate, or build confidence in students.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | EdTech is moderately digitized with growing AI adoption in tutoring platforms, but substitution of human feedback remains limited. Pilots are common; production replacement of human motivational feedback remains rare and cautious. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Education sector adoption of AI tutoring tools is growing but still in early pilot phases relative to sectors like finance or software, with slow institutional uptake. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can scaffold tutors by suggesting encouragement prompts, tracking student confidence signals, and flagging students who need motivation boosts, significantly raising tutor productivity while keeping the human in the feedback loop. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can help tutors draft personalized feedback, suggest encouraging phrasing, and track student progress, meaningfully boosting tutor efficiency while the human remains central to delivery. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate encouragement templates and basic positive feedback, genuine motivational feedback requires understanding each student's emotional state, learning history, and individual psychology. Current systems lack the real-time adaptive judgment and contextual sensitivity needed to replace human feedback reliably at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can generate encouraging feedback text, but genuine relationship-building, reading emotional cues, and adapting motivational strategy in real time to a specific student remains largely beyond current systems for full end-to-end replacement.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Schools and parents often expect human relationships and trust for motivational feedback, and many jurisdictions place educators in a licensed or semi-credentialed role. Organizational preference for human-delivered encouragement and potential liability concerns around AI-only feedback create meaningful resistance. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement blocks AI from giving feedback, but parents/students often prefer human warmth and trust, creating moderate adoption friction rather than hard legal barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-generated encouragement costs near-zero per deployment versus a tutor's hourly rate. Once a system is built, scaling feedback generation has negligible marginal cost compared to human labor. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI feedback generation is cheap per interaction, but achieving comparable motivational quality requires human oversight or hybrid models, narrowing the cost advantage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some AI tutoring products include generic encouragement modules, but deployed systems do not reliably deliver personalized, psychologically-informed positive reinforcement that meaningfully motivates real students. Feedback quality remains material problem in production educational AI. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some AI tutoring products (e.g., Khanmigo, adaptive learning apps) offer canned encouragement and feedback, but they operate narrowly and are not proven substitutes for a human tutor's rapport-building at scale. |
Monitor student performance or assist students in academic environments, such as classrooms, laboratories, or computing centers.
31CI 25–37 · exposure 30 · augmentation 75 · importance 4.1/5 · click for rater detail
Monitor student performance or assist students in academic environments, such as classrooms, laboratories, or computing centers.
31| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Schools and tutoring centers are piloting AI-assisted platforms and analytics, but adoption remains incremental. Widespread deployment has been slowed by budget constraints, technological skepticism in education, and the sector's relatively low digitization compared to finance or professional services. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Education sector adoption of AI is uneven and slower than knowledge-work sectors; digital tools are used for supplemental monitoring but classroom/lab assistance is still predominantly human-delivered. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI tutoring assistants demonstrably enhance tutor productivity through automated grading, progress dashboards, question generation, and instant feedback, keeping the human tutor central to motivation and complex problem-solving. This is a high-augmentation, low-replacement scenario in practice today. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI dashboards and adaptive learning tools significantly help tutors identify at-risk students and personalize support, meaningfully boosting productivity while the tutor remains central. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist with performance tracking and flag learning gaps, but meaningful tutoring requires real-time responsiveness to individual student confusion, emotional state, and adaptive questioning—capacities current systems handle inconsistently. A tutor must gauge comprehension, adjust explanations, and provide encouragement in ways that fall short of 50% time-saving at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can help track quiz scores and flag struggling students, but real-time in-person monitoring, behavioral observation, and hands-on assistance in physical classrooms/labs remain largely human tasks. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Educational institutions face accountability pressure for student outcomes, parental expectations for human interaction, and curriculum standards that often require certified educators or documented instructor oversight. Regulatory compliance and stakeholder acceptance create substantial friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement generally, but institutional preference for human supervision, safeguarding concerns for working with students (especially minors), and need for physical presence create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-driven tutoring platforms have dropped in cost but still require significant setup, content curation, and human oversight. Deployed systems remain comparable to or more expensive than hiring tutors in lower-wage regions, particularly where individualized, adaptive support is demanded. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Software-based performance tracking is cheap, but genuine in-person assistance still requires human staff, making overall cost comparable to or only modestly cheaper than hiring tutors. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Learning management systems and AI tutoring platforms exist in production (e.g., Carnegie Learning, Squirrel AI) and can deliver content and track metrics, but they show material limitations in nuanced student interaction, handling off-topic questions, and making high-stakes pedagogical decisions that justify human tutors. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some ed-tech platforms (adaptive learning systems, LMS analytics) monitor performance data, but comprehensive in-person academic assistance across classrooms and labs is not reliably automated by deployed products. |
Prepare and facilitate tutoring workshops, collaborative projects, or academic support sessions for small groups of students.
30CI 30–30 · exposure 25 · augmentation 75 · importance 3.6/5 · click for rater detail
Prepare and facilitate tutoring workshops, collaborative projects, or academic support sessions for small groups of students.
30| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Educational institutions adopt AI gradually and cautiously—mostly for content generation and administrative support rather than student-facing facilitation. Small tutoring businesses and schools remain heavily human-dependent; production deployment of AI facilitation remains rare. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Education sector adoption of AI for live facilitation is still nascent; pilots for AI tutoring assistants exist but widespread deployment for group facilitation is rare.' |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially assist tutors by generating customized problem sets, pre-session summaries, real-time suggestions for explanations, and post-session feedback synthesis. These augmentations raise tutor productivity and personalization while the educator remains fully in control of live interaction and adaptation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully help tutors prepare lesson plans, generate practice problems, and suggest discussion prompts, boosting prep-time productivity while the human still leads sessions.' |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can draft workshop materials and generate practice problems, but cannot replicate the live facilitation, real-time responsiveness to student confusion, dynamic group management, and adaptive teaching required for effective small-group instruction. The human-interactive core of this task remains largely irreplaceable. |
| Task automatability | claude-sonnet-5 | 2/5 | Preparing materials and generating content for workshops can be assisted by AI, but facilitating live group sessions requires real-time human interaction, classroom management, and adaptive judgment that current AI cannot fully replace end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While there are no strict licensing requirements for tutoring facilitation, parental and institutional preferences for human contact, accountability for learning outcomes, and organizational inertia create moderate friction. Some education contexts have beginning adoption of AI assistance, but wholesale replacement faces trust and liability concerns. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing bars tutoring in most contexts, but institutional preference for human interaction, safeguarding concerns with minors, and need for real-time adaptive facilitation create moderate friction.' |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI content generation is cheap, but end-to-end tutoring still requires human facilitation, oversight, and customization. The cost of AI-generated materials plus human delivery is comparable to or more expensive than direct human tutoring, especially for small groups where personalization matters. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply help draft materials, but the live facilitation component still requires a paid human, so overall cost savings are limited relative to full task replacement.' |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI systems can generate educational content and suggest lesson structures, but no deployed product reliably handles the full facilitation—gauging student understanding, adjusting pacing, managing group dynamics, and providing personalized feedback in real time. Existing tools assist with prep, not live delivery. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tools (e.g., chatbots, content generators) support prep work, but no deployed product reliably facilitates live small-group tutoring sessions in production at scale.' |
Collaborate with students, parents, teachers, school administrators, or counselors to determine student needs, develop tutoring plans, or assess student progress.
28CI 25–30 · exposure 25 · augmentation 75 · importance 4.2/5 · click for rater detail
Collaborate with students, parents, teachers, school administrators, or counselors to determine student needs, develop tutoring plans, or assess student progress.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of AI in tutoring remains in pilot and proof-of-concept phases, concentrated in supplementary content delivery and assessment dashboards rather than in core collaboration and needs-assessment workflows. Most tutoring (especially one-on-one) remains resistant to automation because of the human-relationship and trust requirements. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Education sector adoption of AI for personalized planning is still nascent, with pilots more common than widespread deployment in tutoring and school communication workflows. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can meaningfully augment this task by analyzing student performance data, generating draft tutoring plans and talking points for stakeholder meetings, and producing progress summaries—enabling tutors to spend more time on the irreplaceable work of dialogue, relationship-building, and adaptive judgment with students and families. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully assist tutors by summarizing student data, drafting progress reports, and suggesting tailored resources, boosting efficiency while humans still manage relationships and decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can help generate tutoring plans and provide progress summaries, the core task requires understanding individual student needs through dialogue, building rapport, and collaborative negotiation with multiple stakeholders—activities that demand contextual judgment and interpersonal nuance that current AI systems cannot reliably perform end-to-end. AI tools can assist in data aggregation and draft plan generation, but the human tutor must still conduct the actual needs assessment and stakeholder coordination. |
| Task automatability | claude-sonnet-5 | 2/5 | This task requires multi-party synthesis of relational, contextual, and motivational information that current AI cannot fully gather or interpret without heavy human mediation; only sub-parts (e.g., drafting progress summaries) are automatable. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory and professional barriers are substantial: tutors often work within school systems subject to district oversight; progress assessment and educational planning may require certified educator sign-off; parental consent and trust are prerequisites; and many stakeholders (teachers, counselors) are licensed professionals. Organizations and families typically prefer human judgment and accountability in educational decision-making. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement, but strong parental/institutional preference for human interaction and trust in personalized judgment creates moderate adoption friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The computational cost of AI consultation plus the required human oversight (tutors must still validate needs assessments and coordinate with stakeholders) approaches or exceeds the cost of a tutor handling the task themselves, especially for small-scale or specialized tutoring relationships where personalization is critical. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI can cheaply generate reports or summaries, the human relationship-building and multi-stakeholder coordination still requires paid staff time, keeping overall costs comparable to human-led coordination. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs the full task of needs assessment, plan development, and progress assessment in collaboration with parents, teachers, and administrators. Some educational platforms offer progress tracking and plan templating, but they do not autonomously conduct genuine stakeholder collaboration or adaptive needs assessment at production scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some ed-tech platforms offer AI-generated progress reports or adaptive learning diagnostics, but coordinating with multiple human stakeholders to set tutoring plans is not something deployed products do reliably today. |
Identify, develop, or implement intervention strategies, tutoring plans, or individualized education plans (IEPs) for students.
25CI 25–25 · exposure 25 · augmentation 63 · importance 3.8/5 · click for rater detail
Identify, develop, or implement intervention strategies, tutoring plans, or individualized education plans (IEPs) for students.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Education is a slower-adopting sector with regulatory constraints and high organizational friction. While some schools pilot AI planning tools, production deployment of autonomous IEP development remains rare; most adoption stays at the assisted-drafting level. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Education, especially special education and individualized tutoring, has been slower to adopt AI agents in production compared to sectors like finance or generic office work, with most use still in pilot or assistive stages. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by generating template frameworks, identifying research-backed strategies, or analyzing student data to flag patterns—useful productivity aids for tutors and educators. However, the assistance is limited to structuring or ideation rather than transforming the core diagnostic and relational work. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist tutors by analyzing student performance data, suggesting intervention strategies, and drafting portions of IEPs, significantly speeding up the human's planning process while they retain final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can help draft or suggest tutoring plans and intervention frameworks, the core task requires understanding a specific student's cognitive, emotional, and learning profile through assessment and interaction—work that depends on human judgment of nuance and context. Current systems cannot autonomously develop effective, legally compliant IEPs that meet the ≥50% time-saving threshold while maintaining quality. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can help draft or suggest components of an IEP or tutoring plan, but developing and implementing an appropriate individualized plan requires ongoing assessment of a specific student, professional judgment, and legal compliance that AI cannot fully replace today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | IEPs are legally mandated documents in special education; they must be signed by licensed educators and parents. Federal law (IDEA) requires human professional judgment and multi-disciplinary team sign-off, creating strong regulatory and liability barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | IEPs are often legally mandated documents requiring certified professionals (special education teachers, psychologists) to develop and sign off, creating strong regulatory and liability barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tutoring and planning tools reduce some costs but require continuous human review, customization, and accountability—especially for legal IEP compliance. The all-in cost of AI-assisted planning with required human oversight is comparable to or higher than direct tutor labor for this specialized task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools can lower drafting costs for portions of documentation, but the ongoing individualized assessment, monitoring, and implementation still require substantial paid human tutor/specialist time, keeping overall cost comparable to human-driven work. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some educational platforms offer AI-assisted lesson planning and template-based IEP generators, but no mature product reliably performs end-to-end IEP development or intervention-strategy design in production without substantial human oversight. These remain largely advisory tools rather than autonomous performers. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some ed-tech products offer adaptive learning suggestions or IEP drafting aids, but no deployed product reliably identifies, develops, and implements a full individualized education plan without significant human oversight. |
Organize tutoring environment to promote productivity and learning.
13CI 5–21 · exposure 0 · augmentation 25 · importance 3.9/5 · click for rater detail
Organize tutoring environment to promote productivity and learning.
13| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Tutoring remains a low-digitization, high-human-contact service sector. Physical environment organization has not seen measurable AI adoption, and tutoring work is concentrated in small practices and one-on-one settings with limited infrastructure for automation. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Tutoring and education sectors show moderate, uneven AI adoption mostly for content delivery, not for environmental/session organization, which lags. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by suggesting evidence-based learning environment designs or analyzing data on learner preferences, but most of the task (physical arrangement, real-time adjustment, interpersonal calibration) requires human execution with limited AI enhancement potential. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can suggest session structure, materials, or scheduling tips, offering minor planning assistance, but does not meaningfully enhance the physical/environmental organization task itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Organizing a tutoring environment requires real-time physical arrangement of spaces, assessment of individual learner needs, and adaptive environmental design—tasks that demand embodied presence and human judgment about learning conditions. Current AI cannot physically organize environments or make contextual decisions about optimal learning setups. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physically arranging a real-world space or setting up a session context to suit an individual learner's needs, which is a physical/logistical judgment task AI cannot perform end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Organizing a learning environment is inherently tied to direct human-learner interaction and the tutor's professional responsibility for creating an effective pedagogical context. Educational roles typically require human presence, and liability for learning outcomes rests with the human educator. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing barrier exists, but the task's inherently physical and relational nature (arranging space, building rapport) creates practical friction against automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task involves physical space setup, personal interaction with learners, and real-time troubleshooting—activities where human labor is either necessary or cheaper than the overhead of deploying robotics and AI systems to accomplish equivalent results. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Since AI cannot substitute for the physical/organizational aspects, any AI cost is additive rather than replacing the human's labor here, making cost comparison unfavorable to AI substitution. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably organizes physical tutoring environments end-to-end. While AI can suggest organizational principles or learning environment design, it cannot independently execute the physical and interpersonal coordination required. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product organizes physical or session environments for tutoring; this remains a human logistical and interpersonal task. |
Travel to students' homes, libraries, or schools to conduct tutoring sessions.
7CI 0–15 · exposure 0 · augmentation 13 · importance 3.3/5 · click for rater detail
Travel to students' homes, libraries, or schools to conduct tutoring sessions.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No meaningful displacement is occurring in this task; even as remote tutoring and AI tutoring assistants exist, the specific task of traveling to conduct in-person sessions remains performed by human tutors. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Physical travel tasks in education services show negligible AI adoption since there is no robotic or virtual equivalent being deployed at scale. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While AI could marginally assist tutors in preparing materials or planning sessions before traveling, it offers minimal assistance with the core task of traveling to and conducting sessions at specific locations. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI cannot assist with the physical act of traveling to a location; navigation apps offer trivial route assistance but do not meaningfully change this task's productivity. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical presence at a specific location (student's home, library, or school), which AI systems cannot do. Travel itself is entirely non-automatable by current AI. |
| Task automatability | claude-sonnet-5 | 1/5 | Physical travel to a location cannot be performed by current AI systems, which have no embodied presence; this is purely a physical logistics component of the task.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong barriers exist: tutoring at a specific physical location is inherently a human-contact requirement, and parents/students typically expect in-person interaction for educational services. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing barrier specifically blocks automation, but the inherent physical/embodiment requirement and in-person trust with minors create practical friction beyond simple software substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI cannot perform this task, making cost comparison irrelevant. The task fundamentally requires human mobility and presence. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute for physical travel, so cost comparison is inapplicable/AI is not a viable cheaper alternative for this specific action. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can travel to physical locations or conduct in-person tutoring sessions. This task remains firmly in the domain of human execution. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product physically transports itself to a home, library, or school; this remains entirely outside current AI capability. |
Related occupations — Educational Instruction & Library
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.