Social and Human Service Assistants
21-1093.00Median wage $45,930/yr437,860 employed (US)Rank #418 of 923 scored · top 45% by substitution
Assist other social and human service providers in providing client services in a wide variety of fields, such as psychology, rehabilitation, or social work, including support for families. May assist clients in identifying and obtaining available benefits and social and community services. May assist social workers with developing, organizing, and conducting programs to prevent and resolve problems relevant to substance abuse, human relationships, rehabilitation, or dependent care.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution29
Exposure26
Augmentation56
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
5%
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%28
panel mean rating 2.1/5 → substitution pressure 28/100
Technical feasibility todayw 20%24
panel mean rating 2.0/5 → substitution pressure 24/100
Cost vs. human wagew 15%36
panel mean rating 2.4/5 → substitution pressure 36/100
Adoption barriersw 20%inverted — strong barriers lower the score37
panel mean rating 3.5/5 (barrier strength) → substitution pressure 37/100
Sector adoption velocityw 10%18
panel mean rating 1.7/5 → substitution pressure 18/100
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.
Inform tenants of facilities, such as laundries or playgrounds.
79CI 74–84 · exposure 75 · augmentation 50 · importance 2.6/5 · click for rater detail
Inform tenants of facilities, such as laundries or playgrounds.
79| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Property management and tenant communication are increasingly digital sectors; many residential facilities already use automated systems for routine notifications, with widespread adoption in mid-to-large properties. Momentum is clearly toward automation in this space. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social and human services is a sector with historically low digitization and slow AI adoption, though property management software adoption is somewhat further along. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted drafting of facility announcements or multi-channel scheduling tools can help social service assistants compose and distribute information more efficiently, though the task itself is simple enough that augmentation impact is moderate rather than transformative. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-driven communication tools can draft notices, automate reminders, and answer routine facility questions, freeing staff for more complex interpersonal tasks. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Informing tenants of facility availability and features is straightforward factual communication that AI can automate nearly completely via email, SMS, or in-app notifications. Setup requires facility information input, but once configured, the system requires minimal ongoing human intervention to deliver consistent, timely information to residents. |
| Task automatability | claude-sonnet-5 | 4/5 | This is a simple, repetitive information-provision task (notifying tenants about amenities) that can be handled via automated messaging, chatbots, or notice systems with minimal quality loss. |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | Few legal or regulatory barriers exist to automating facility information distribution. Some tenant preference for human interaction may exist, and organizations may choose to retain human touch, but nothing legally requires a human to perform this task, and oversight burden is minimal. |
| Adoption barriers | claude-sonnet-5 | 1/5 | There is no licensing, liability, or regulatory requirement that a human must personally inform tenants of facility availability. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated messaging and notification platforms cost pennies per tenant per month, while a social service assistant manually informing tenants (phone calls, in-person, individual emails) would cost significantly more in labor. The AI cost is orders of magnitude cheaper. |
| Cost vs. human wage | claude-sonnet-5 | 5/5 | Automated messaging/notification systems cost pennies per tenant compared to staff time spent individually informing tenants. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Deployed systems (automated messaging platforms, tenant portals, property management software with notification features) reliably send facility information at scale today. Minor limitations exist in handling unusual edge cases or personalized preferences, but production-grade systems handle routine tenant notifications with high reliability. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Tenant communication platforms, property management software, and chatbots already deploy automated notifications and FAQ responses about building facilities in production today. |
Assist in planning food budgets, using charts or sample budgets.
56CI 46–65 · exposure 58 · augmentation 75 · importance 3.1/5 · click for rater detail
Assist in planning food budgets, using charts or sample budgets.
56| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social services remain among the least digitized and slowest-adopting sectors for AI; budget planning tools exist but are rarely AI-driven, and organizational culture emphasizes human-centered service delivery and client relationships over efficiency automation. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a historically slow-adopting, resource-constrained sector with limited AI tool integration in daily casework despite general availability of budgeting software. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially boost human productivity by auto-generating draft budgets, creating visual comparisons, and offering instant scenario modeling, allowing the caseworker to focus on dialogue, contextual understanding, and client support rather than manual calculation and chart creation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can quickly draft, adjust, and explain sample food budgets, letting human assistants focus on personalized counseling and follow-up, meaningfully boosting efficiency. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can generate charts, sample budgets, and cost-benefit comparisons from existing templates fairly quickly, but meaningful planning requires understanding individual circumstances, nutrition needs, and local prices that typically demand human interaction and iteration. |
| Task automatability | claude-sonnet-5 | 4/5 | Budget planning with charts/templates is a structured, quantitative task that current AI (spreadsheets with LLM assistance, chatbots) can largely automate, generating sample budgets and calculations quickly. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Social service delivery typically requires licensed or credentialed staff to engage with clients; liability concerns around financial advice, regulatory requirements around service delivery, and strong organizational reliance on human relationship-building create meaningful friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement to create a budget plan, though case workers may need to verify accuracy and maintain client trust/personal relationship, creating mild friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference and integration costs for budget generation are low, and per-task cost is substantially below the loaded wage of a social service assistant, though oversight and refinement add some overhead. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Generating or adjusting a food budget via AI tools costs a fraction of a human caseworker's time, though some oversight and personalization still add cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Existing AI systems can produce budget templates and charts reliably, but production deployments in social service organizations remain limited; most real-world adoption still relies on human caseworkers with spreadsheet or online tools rather than AI-driven planning systems. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Financial planning assistants and budgeting apps exist and are used in production, but few are tailored specifically to social-service food budgeting with client-specific constraints, so reliability in this niche context is moderate. |
Assist clients with preparation of forms, such as tax or rent forms.
51CI 46–56 · exposure 50 · augmentation 75 · importance 3.0/5 · click for rater detail
Assist clients with preparation of forms, such as tax or rent forms.
51| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social service organizations are typically lower-digitization, under-resourced sectors with slow AI adoption. While some government and nonprofit agencies pilot form automation, production deployment at scale remains uncommon and spotty. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a historically slow-adopting, underfunded sector with limited IT infrastructure, so despite the availability of AI tools, actual production use in this occupation remains limited. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI form assistants can significantly augment human caseworkers by pre-filling data, flagging errors, suggesting correct fields, and generating drafts—allowing workers to focus on eligibility determination and client counseling rather than manual data entry. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully speed up form preparation by pre-filling fields, explaining requirements, and catching errors, while the human assistant remains in the loop for client interaction and judgment calls. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can automate parts of form preparation (data extraction, field population, basic guidance) but typically requires human review for accuracy, legal liability, and client-specific context verification. The task involves both routine data entry and judgment about eligibility and correctness. |
| Task automatability | claude-sonnet-5 | 3/5 | Form-filling based on client-provided information is a structured, language-based task that current AI can largely handle, but understanding client circumstances and ensuring accurate completion often requires clarifying dialogue and empathy that reduce full automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Significant barriers exist: tax and benefits form completion can carry legal/liability implications, clients often require in-person assistance or interpreter services, and organizational reliance on human caseworkers creates institutional friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No formal licensing is required for this specific task, but agencies serving vulnerable clients often require human staff for trust, accessibility, and liability reasons, creating moderate organizational friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-driven form assistance is substantially cheaper than human labor once deployed; however, integration costs and compliance review overhead partially offset savings. The per-task cost is likely 2–5× cheaper than a human assistant. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI-assisted form tools cost a fraction of a human assistant's time per form, though some human oversight is still needed to catch errors or handle edge cases, slightly reducing the cost advantage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Document automation and form-filling tools exist in production (e.g., tax software, benefits application systems) but often require human oversight and frequently have gaps in handling complex or edge-case scenarios common in social services. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Tax software and AI chatbots already assist with form preparation in production (e.g., TurboTax, benefits navigators), but for vulnerable populations needing hands-on guidance, deployed AI-only solutions remain narrow and error-prone. |
Explain rules established by owner or management, such as sanitation or maintenance requirements or parking regulations.
49CI 32–65 · exposure 45 · augmentation 75 · importance 3.0/5 · click for rater detail
Explain rules established by owner or management, such as sanitation or maintenance requirements or parking regulations.
49| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social service organizations are generally lower-digitization, resource-constrained sectors with slower AI adoption; while rule-documentation tools exist, production deployment of AI for client-facing explanation remains minimal in this occupational space. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social/human services and property management sectors have historically low digitization and slower AI adoption compared to finance or tech, though tenant-facing chatbots are slowly spreading. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can meaningfully assist by drafting rule explanations, generating multilingual versions, creating FAQs, and helping social workers prepare personalized explanations for specific clients, thereby raising staff productivity while the human retains judgment on delivery and context. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can draft explanations, generate FAQ responses, and provide quick lookup support for human service assistants explaining rules, improving consistency and speed while humans retain oversight for sensitive cases. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task involves explaining rules orally or in writing, which LLMs can draft, but the context-dependent, personalized nature of rule explanation to diverse residents/clients and the requirement to adjust tone and complexity based on audience comprehension makes full end-to-end automation unlikely to meet the 50% time-saving threshold consistently. |
| Task automatability | claude-sonnet-5 | 4/5 | Explaining established rules is largely an information-delivery task that AI chatbots/voice agents can handle end-to-end for most routine cases, given the rules are documented and static.: |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Social and human service work typically requires licensed staff (social workers, counselors) or employees with specific certifications to interact with vulnerable populations; liability concerns and regulatory requirements around who may represent the organization's policies to clients create material legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement to explain rules, but some organizational preference for human contact in resident/tenant relations creates mild friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI can generate rule documentation and explanations at minimal cost per instance; however, ongoing oversight and personalization for specific contexts still requires some human involvement, keeping the cost advantage significant but not orders of magnitude. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | A chatbot or automated phone system answering rule questions costs a small fraction of staff time once configured, though initial setup and periodic updates add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can generate rule explanations and FAQs, deployed products rarely reliably handle the nuanced, interpersonal aspect of explaining rules to individuals with varying literacy, language, and emotional states in real social service settings at production scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Property management and community chatbots exist and handle FAQ-style rule explanations, but reliability drops for nuanced or edge-case tenant questions, so scope is narrow. |
Advise clients regarding food stamps, child care, food, money management, sanitation, or housekeeping.
47CI 34–60 · exposure 45 · augmentation 75 · importance 3.4/5 · click for rater detail
Advise clients regarding food stamps, child care, food, money management, sanitation, or housekeeping.
47| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social service agencies and nonprofits are generally slower to adopt automation than private tech and finance sectors, often due to underfunding, legacy systems, regulatory caution, and cultural preference for human-centered care. Pilots exist, but deep production deployment remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a historically low-digitization, underfunded sector with slow technology adoption despite some emerging chatbot pilots for benefits navigation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially assist caseworkers by auto-generating eligibility summaries, drafting benefit application guidance, and surfacing relevant resources, freeing staff to focus on relationship-building and complex problem-solving. This is already proving effective in practice as an assistive layer rather than full replacement. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by quickly summarizing eligibility rules, drafting resource lists, and answering routine client questions, freeing case workers for higher-touch counseling. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Most client advice on food stamps, child care subsidies, money management basics, and sanitation can be systematized into rule-based recommendations or retrieval-augmented responses. Current AI can draft personalized guidance, check eligibility rules, and provide structured advice at substantial time savings, though complex case judgment and sensitive rapport-building would still benefit from human review. |
| Task automatability | claude-sonnet-5 | 2/5 | While AI chatbots can provide general information on food stamps or budgeting, effective advising requires assessing individual circumstances, eligibility nuances, and emotional context that current systems cannot reliably handle end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While no law mandates a licensed human must deliver basic advisory services, organizational accountability for wrong advice, regulatory compliance in benefits systems, and client vulnerability create meaningful friction. Many agencies prefer human sign-off and face community/management resistance to full automation of sensitive services. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement, but case management often involves documentation, liability concerns, and client trust that favor human involvement, plus agency policies requiring human sign-off. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | After initial setup, inference cost for AI-generated advice is negligible compared to the loaded wage of a human social services assistant. Oversight and occasional human intervention add cost, but remain well below the hourly cost of staff, yielding a significant per-task advantage. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI tools for information lookup are cheap, but the need for human verification, follow-up, and personalized guidance keeps overall cost comparable to human labor for full task completion. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Chatbots and AI advisory systems exist in production for benefits eligibility and financial guidance, but they typically handle narrow domains (e.g., tax software) or have material accuracy gaps in multi-benefit scenarios. Social service agencies have begun piloting AI advisors, but few scale them without human oversight due to error costs and client trust concerns. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some benefits-navigation chatbots and eligibility screeners exist, but they are narrow tools supplementing rather than replacing human advisors for holistic case management. |
Provide information or refer individuals to public or private agencies or community services for assistance.
42CI 28–56 · exposure 38 · augmentation 75 · importance 3.6/5 · click for rater detail
Provide information or refer individuals to public or private agencies or community services for assistance.
42| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Social services organizations tend toward low digitization, resource constraints, and reliance on human relationships; adoption of AI for case referral remains nascent and confined largely to pilots or supplementary tools rather than production replacement. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a historically under-digitized, resource-constrained sector with slow technology adoption relative to finance or professional services. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially assist by rapidly searching agency databases, generating candidate referral lists, drafting summaries of services, and flagging options based on client profile—freeing the human assistant to focus on relationship-building, nuanced assessment, and complex problem-solving while human judgment remains in the loop. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can efficiently surface relevant agencies, eligibility criteria, and contact details, significantly speeding up the assistant's referral research while the human manages the client relationship. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Providing referrals requires understanding client needs, matching them to appropriate services, and communicating options—tasks with significant human judgment and context-dependency. While AI can search databases and generate referral lists, the assessment of individual circumstances and the nuanced judgment needed to match people to services remains largely manual and difficult to automate end-to-end with equal quality. |
| Task automatability | claude-sonnet-5 | 3/5 | AI chatbots can already match clients to resources and generate referral information from structured databases, but nuanced eligibility assessment and follow-up require human judgment, limiting full automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | These roles often operate in regulated social services contexts where human accountability, ethical judgment, and documented decision-making by qualified staff are expected or mandated; while not absolute barriers, liability concerns and organizational preference for human touch create meaningful friction to automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing typically required for basic information/referral, though some contexts (e.g., mental health, child welfare) impose oversight and liability concerns that favor human involvement. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | An AI system with sufficient accuracy and integration would need substantial oversight and validation by trained staff; the total cost (AI inference, integration, human oversight, liability) remains comparable to or potentially higher than the cost of a human assistant performing the task directly. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated lookup and referral generation is extremely cheap compared to a human assistant's time, especially for straightforward informational requests. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No mature deployed system reliably handles the full scope of needs assessment and personalized referral matching at scale; while AI can retrieve agency information and draft referral suggestions, real-world deployments still require human case workers to validate recommendations and customize responses to individual circumstances. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | 211-style resource navigation chatbots and referral tools exist in production, but they often have narrow scope and require human verification for complex or sensitive cases. |
Keep records or prepare reports for owner or management concerning visits with clients.
41CI 34–48 · exposure 45 · augmentation 75 · importance 4.0/5 · click for rater detail
Keep records or prepare reports for owner or management concerning visits with clients.
41| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social service organizations are among the slowest to adopt automation, given regulatory constraints, small budgets, and institutional conservatism around client records. Pilot projects exist but production deployment remains limited and cautious. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services agencies are generally under-resourced and slower to adopt AI tools compared to finance or tech sectors, with documentation automation only slowly appearing in case management software. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can substantially assist by auto-populating forms, summarizing notes from visit recordings, flagging missing required fields, and generating draft reports—freeing caseworkers from administrative drudgery while they retain oversight and responsibility for accuracy. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully speed up drafting, summarizing, and organizing visit notes and reports, letting human service assistants focus more time on client interaction while still reviewing and finalizing records themselves. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can partially automate record-keeping and report generation from structured intake data or conversation transcripts, but requires significant human review and quality assurance due to sensitivity of client information and need for accurate, compliant documentation. The task typically involves judgment about what details matter and how to frame them, which slows full automation. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft summaries and organize case notes from dictated or typed input, but the underlying client interaction and judgment about what to record still require human involvement, capping full end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong legal and regulatory barriers exist: HIPAA compliance, state licensing rules for social workers, client confidentiality requirements, and organizational liability if AI-generated records are inaccurate or missing critical details. Many jurisdictions require a licensed professional to sign off on case notes, creating a legal gate. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Case records often carry confidentiality obligations, agency compliance rules, and require staff sign-off for accuracy and liability, creating moderate barriers even though no explicit licensure mandates human authorship of the notes. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While AI-assisted documentation is cheaper per unit than hiring administrative staff, the oversight required to ensure accuracy and compliance (legal liability if records are wrong) increases total cost. Full automation is not yet cost-effective enough to clearly undercut human record-keeping. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Using AI to draft records could cut documentation time significantly, but integration with case management systems, privacy safeguards, and required human review keep costs from being drastically lower than the marginal cost of a human doing it directly. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Document management and basic report generation products exist and are deployed in some social service organizations, but they struggle with the nuance required for sensitive client information, clinical accuracy, and regulatory compliance. Most implementations require substantial human review and editing. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | General-purpose AI writing tools and some EHR/case-management systems offer note-drafting or summarization features, but few deployed products reliably generate compliant, accurate client visit reports in social services settings at scale. |
Interview individuals or family members to compile information on social, educational, criminal, institutional, or drug history.
36CI 11–60 · exposure 33 · augmentation 63 · importance 3.8/5 · click for rater detail
Interview individuals or family members to compile information on social, educational, criminal, institutional, or drug history.
36| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social services remain heavily human-centered, under-digitized, and fragmented across public agencies with slow IT adoption; pilot automation exists but deployment in production is lagging relative to higher-velocity sectors. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a historically low-digitization, human-contact-intensive sector with slow AI adoption for direct client interviewing tasks, though administrative AI tools are creeping in. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can powerfully assist a human interviewer by transcribing conversations, flagging missing fields, summarizing prior records, and suggesting follow-up questions—keeping the human in control while significantly raising data quality and completeness. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help by transcribing, summarizing, or pre-structuring interview notes and flagging follow-up questions, providing moderate assistance while the human conducts the actual interview. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Structured interviews to extract and compile historical information can be substantially automated using AI-assisted forms, document analysis, and information extraction; once basic facts are gathered (dates, institutions, charges), manual review for nuance or sensitive details adds minimal time, meeting the ≥50% savings bar. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires building rapport and trust with vulnerable individuals to elicit sensitive personal history, which demands human judgment, empathy, and adaptive follow-up questioning that current AI cannot reliably replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Interview subjects may have legal rights to human contact and privacy considerations in criminal/institutional history contexts; organizational norms and client trust also favor human presence, though these are social rather than hard legal barriers. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Confidentiality, mandated-reporter obligations, trust-building with vulnerable clients, and agency/legal requirements around case documentation create strong barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI inference and integration costs for automated interview transcription and information extraction are substantially lower than the loaded wage of a social service assistant, especially when oversight is light; only significant human review adds cost. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Even if AI-assisted transcription or intake forms exist, the core interviewing and judgment work still requires a skilled human, so all-in cost savings versus the human interviewer are minimal. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed chatbot and form-filling systems exist and can conduct basic scripted interviews, but real-world reliability remains uneven due to follow-up probing, detecting evasion, and handling emotional disclosure—production use remains limited to data extraction stages rather than full replacement. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product conducts sensitive intake interviews with vulnerable populations (e.g., drug history, criminal history) autonomously in production; this remains outside current AI product scope. |
Submit reports and review reports or problems with superior.
33CI 30–35 · exposure 30 · augmentation 75 · importance 3.9/5 · click for rater detail
Submit reports and review reports or problems with superior.
33| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social and human service sectors have slower digital maturity and lower AI adoption rates than information or finance; these are often smaller, non-profit, or government organizations with limited resources and conservative practices around case management communication. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a sector with historically slower AI adoption due to funding constraints, regulatory environment, and reliance on human judgment in case management. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can meaningfully assist by drafting templates, summarizing case notes, flagging anomalies, and organizing information for review, allowing the assistant to focus on judgment and communication with their supervisor rather than rote data entry and formatting. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by drafting report summaries, organizing case notes, and flagging issues for discussion, improving efficiency while the human retains the supervisory interaction. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can draft reports and flag issues for review, the task fundamentally requires human judgment about what constitutes a 'problem' worth escalating and how to frame issues for a superior. Current systems cannot reliably autonomously decide whether to escalate or how to communicate context-sensitive concerns, limiting time savings to under 50%. |
| Task automatability | claude-sonnet-5 | 2/5 | Report submission can be templated and drafted with AI assistance, but the discussion of problems with a superior requires real-time judgment, context, and relational communication that current AI cannot fully replace end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | Social service organizations often have documentation and compliance requirements that create oversight friction, and there is organizational preference for human judgment in escalation decisions. However, no strict licensing requirement prevents a human from delegating report assembly to AI. |
| Adoption barriers | claude-sonnet-5 | 3/5 | There's no strict licensing requirement for this specific task, but organizational accountability structures and supervisory oversight norms create moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI can reduce clerical work in report drafting, but deployment requires integration with organizational systems, quality oversight, and potential rework when judgment fails. For a task involving judgment and organizational communication, setup and oversight costs partially offset savings versus a modestly-paid assistant. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply draft report text, but the human-to-human review and problem discussion component still requires paid staff time, keeping overall cost savings modest. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Document generation and basic report assembly tools exist in production (e.g., automated form-filling, templated report generation), but the critical review and judgment component—assessing which problems merit escalation—remains inconsistent and error-prone without human oversight. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Products like document generators and summarization tools exist to help draft reports, but no deployed system autonomously reviews case problems with a supervisor in a reliable, production-grade way for this role. |
Observe clients' food selections and recommend alternate economical and nutritional food choices.
31CI 25–37 · exposure 20 · augmentation 63 · importance 2.9/5 · click for rater detail
Observe clients' food selections and recommend alternate economical and nutritional food choices.
31| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social services remain a largely human-intensive, lower-digitization sector with slow AI adoption. Most social service organizations still rely on direct staff observation and counseling; pilot automation programs are rare and organizational resistance to removing human contact in client assessment is substantial. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a sector with generally low AI adoption and digitization compared to finance or tech, with pilots for chatbots existing but direct client observation tasks rarely touched by AI in production. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can meaningfully assist social service workers by instantly cross-referencing nutrition databases, generating cost-benefit analyses of food swaps, and flagging dietary concerns, significantly reducing research time and improving suggestion quality while the worker maintains judgment and interpersonal contact with the client. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help human service assistants by generating nutritional information, budget-friendly meal suggestions, or educational materials to support their recommendations, improving efficiency in the advisory portion of the task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can analyze food selections against nutritional data and suggest alternatives, but the task requires observing actual client behavior in real time and making personalized recommendations that account for cultural preferences, allergies, and budget constraints that are not always explicitly stated. Current AI lacks reliable real-world behavioral observation and contextual understanding needed for consistent, trusted guidance. |
| Task automatability | claude-sonnet-5 | 2/5 | This task requires real-time in-person observation of a client's behavior and personalized, context-sensitive counseling that current AI cannot autonomously perform end-to-end.dynamic w |
| Adoption barriers | claude-haiku-4-5-20251001 | 3/5 | While there is no strict licensing requirement for nutritional guidance by assistants (unlike registered dietitians), organizational practice, liability concerns around dietary recommendations, and strong institutional preference for human rapport and accountability in client services create moderate friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement for this specific task, but it occurs within social service delivery where client trust, rapport, and case-context sensitivity create real friction against pure automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-based nutrition recommendation systems have minimal marginal cost per interaction compared to the loaded wage of a human social service assistant conducting the same analysis and counseling, making automation economically favorable if feasibility were solved. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Since AI cannot independently perform the observation and personalized counseling, any AI use still requires substantial human involvement, keeping combined costs close to or above the human-only cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While nutrition analysis tools and chatbots exist, deployed products have not demonstrated reliable real-world performance at the observational and interpersonal level required. Systems can generate generic nutritional suggestions but struggle with the nuanced assessment of individual client circumstances that this task demands. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product observes clients' actual food selections in situ and delivers tailored nutritional/economic recommendations as part of a human-service interaction; this remains research-adjacent at best. |
Assist in locating housing for displaced individuals.
28CI 25–30 · exposure 25 · augmentation 50 · importance 3.8/5 · click for rater detail
Assist in locating housing for displaced individuals.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social service agencies are slow to digitize, operate under funding constraints, and maintain human-centered workflows; while some may use basic search tools, deep AI adoption for this task remains limited and primarily pilot-stage in most organizations. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a sector with historically low AI adoption and digitization, with pilots for case management tools emerging slowly and unevenly across agencies. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by generating housing lists, flagging affordability mismatches, and organizing search results, meaningfully reducing manual research time and improving options presented to caseworkers without removing human judgment from selection and placement. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help assistants search housing databases, draft communications, and track available units, meaningfully speeding up parts of the task even though the human remains central to placement decisions. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can help search listings and match criteria, the task requires human judgment in assessing client needs, evaluating suitability, and negotiating with landlords—often involving sensitive circumstances and interpersonal factors that cannot be fully automated to the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | Searching listings or databases could be AI-assisted, but assessing individual client needs, negotiating with landlords, and coordinating in-person logistics require human judgment and relationship-building that current AI cannot replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: social service provision often requires licensed caseworkers or certified assistants for displaced-person programs; liability for poor housing recommendations is high; and many clients need in-person relationship-building that regulations and organizational policy reserve for humans. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement mandates a human specifically, but housing placement often involves compliance with social service regulations, trust-building with vulnerable populations, and coordination with third parties like landlords and shelters. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for property search are inexpensive, but integrating oversight, case management context, and human follow-up means total cost approaches or exceeds the loaded wage of a human assistant performing this task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools could cheaply assist with search and database queries, but the bulk of the task still requires human caseworker time for outreach, negotiation, and follow-through, keeping overall costs comparable to human labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some AI tools exist for property search and matching (Zillow-like systems), but no mature deployed product reliably handles the full task of locating suitable housing for displaced individuals, which demands contextual understanding of affordability, accessibility, and crisis coordination. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some housing-search and eligibility-matching tools exist in social services, but no deployed product reliably handles the full process of locating and securing housing for displaced individuals. |
Develop and implement behavioral management and care plans for clients.
24CI 23–25 · exposure 25 · augmentation 63 · importance 4.1/5 · click for rater detail
Develop and implement behavioral management and care plans for clients.
24| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Social services remains a labor-intensive, often under-resourced sector with limited digitization; adoption of AI for core clinical tasks lags far behind high-tech sectors, with most organizations still using basic documentation systems and prioritizing human relationship-building. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a historically low-digitization, human-contact-intensive sector with slow, cautious AI adoption compared to finance or tech. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist by offering evidence-based intervention suggestions, organizing client data, or drafting template sections, thereby reducing documentation burden and helping practitioners consider broader options—but the human service provider remains essential for individualization, client engagement, and adaptive implementation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by drafting plan templates, summarizing client history, suggesting evidence-based interventions, and reducing documentation burden, while the human retains judgment and implementation responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can help draft care plans by synthesizing client information and suggesting standard interventions, developing and implementing behavioral plans requires ongoing assessment of individual client responses, clinical judgment about plan modifications, and real-time behavioral observation—tasks that current AI systems cannot execute end-to-end with 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | Care plan development requires nuanced clinical judgment, client rapport, and contextual knowledge that current AI cannot reliably synthesize end-to-end, though it can draft templates or summarize notes. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Most jurisdictions require licensed social workers, counselors, or behavioral health professionals to develop and oversee care plans; liability for client harm from faulty plans is high and legally assigned to the human provider, and many clients require in-person human contact for effective behavioral intervention. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Care plans often require credentialed professional sign-off, adherence to regulatory/ethical standards, and accountability for client welfare, creating strong barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Implementation of behavioral plans demands constant human presence, observation, and adjustment; while AI might reduce plan-drafting costs marginally, the labor-intensive nature of client interaction and real-time behavioral management means AI cannot achieve cost parity, let alone significant savings, for the full task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI drafting tools are cheap per query, but the need for extensive human review, client interaction, and liability oversight keeps overall cost comparable to or only modestly cheaper than human labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some AI tools exist to assist with documentation and plan templating, but no deployed product reliably develops and implements individualized behavioral management plans independently; the task requires continuous human judgment, client interaction, and adaptive intervention that deployed systems do not perform at production scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some case management software includes AI-assisted documentation or suggestion features, but no deployed product independently creates and implements individualized behavioral care plans reliably in production. |
Observe and discuss meal preparation and suggest alternate methods of food preparation.
20CI 5–35 · exposure 13 · augmentation 38 · importance 2.9/5 · click for rater detail
Observe and discuss meal preparation and suggest alternate methods of food preparation.
20| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Social and human service sectors are among the slowest to digitize and adopt AI automation; this task involves vulnerable populations and interpersonal trust, so adoption remains negligible in production settings. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Social and human services work is low-digitization, relationship-based, and physically situated, showing minimal AI adoption for hands-on client observation tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could usefully assist by generating meal preparation suggestions and alternatives for a human worker to present, or by flagging safety concerns in video footage, but the core task of observation and discussion requires human judgment and presence. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could supply recipe suggestions or nutrition information the assistant relays, but it cannot meaningfully augment the observational and interpersonal core of the task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can analyze meal preparation techniques via video or image analysis and suggest alternatives, the task fundamentally requires real-time observation of a person's cooking, direct interaction, and adaptive discussion tailored to individual needs—capabilities current systems lack at scale. End-to-end automation with 50% time savings at equal quality is not demonstrable today. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires in-person observation of a client's physical meal preparation and real-time interpersonal coaching, which current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | High barriers exist: the task occurs in sensitive care contexts (elderly, disabled, at-risk populations), where human judgment, liability concerns, and organizational preference for licensed staff oversight make direct human service legally and professionally expected in most jurisdictions. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not formally licensed in most jurisdictions, this task involves direct client contact, trust-building, and safety judgment that organizations and clients expect from a human helper. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | AI-powered video analysis and recommendation systems are inexpensive to deploy per interaction compared to a trained human social service assistant's loaded wage, though integration and personalized interaction still require overhead. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot perform the physical observation and coaching component, there is no viable AI-based substitute cost to compare against the human wage. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Computer vision systems can identify food items and cooking methods, but deployed products struggle with real-time, contextual assessment of meal preparation nuances and lack the interactive discussion component. No mature production system reliably observes and engages in adaptive dialogue about cooking alternatives. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product observes clients cooking and provides embodied, contextual feedback; this remains outside current product capability. |
Assess clients' cognitive abilities and physical and emotional needs to determine appropriate interventions.
16CI 7–25 · exposure 13 · augmentation 50 · importance 4.4/5 · click for rater detail
Assess clients' cognitive abilities and physical and emotional needs to determine appropriate interventions.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Social service agencies are predominantly small, under-resourced, and slow to digitize; while some adopt digital intake and screening forms, genuine AI-driven assessment replacement is minimal and adoption remains in pilot stage rather than widespread production deployment. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a historically low-digitization, human-contact-intensive sector with slow AI adoption for direct client assessment work, though administrative AI tools are creeping in. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by organizing client information, flagging potential risk factors, generating preliminary intake summaries, or suggesting assessment frameworks, thereby supporting human assessors' efficiency without replacing clinical judgment. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help assistants organize intake notes, flag risk indicators from documented history, or suggest resources, but the core assessment interaction remains human-led. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can gather and process some information about client needs through questionnaires or intake forms, comprehensive assessment of cognitive, physical, and emotional needs requires nuanced human judgment, observation of nonverbal cues, and relational rapport that current AI cannot reliably replicate end-to-end at equal quality. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires in-person observation, rapport-building, and clinical judgment about vulnerable individuals' physical, cognitive, and emotional states—AI cannot reliably perceive or synthesize this holistically today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Assessment of emotional and cognitive needs often requires licensed social workers, psychologists, or certified healthcare providers to conduct and sign off on recommendations; liability, regulatory requirements (e.g., HIPAA, state licensure), and client welfare create strong legal and professional barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Many jurisdictions require credentialed human judgment for needs assessments tied to service eligibility, safety, and liability, and clients often need direct human interaction for trust and disclosure. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Even where AI-assisted screening tools exist, they typically require human expert review and validation, so the combined cost remains comparable to or higher than direct human assessment, especially when liability and accuracy requirements are factored in. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot perform the core assessment task, so any AI substitute would still require a human assessor, making AI-only delivery not cheaper but simply infeasible for the core function. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some screening tools and intake questionnaires exist as digital products, but they are narrow adjuncts to professional assessment rather than replacements; clinical-grade cognitive and emotional assessment remains dependent on trained human professionals in production settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently assesses client needs and determines interventions in social services; this remains a research-stage aspiration at best, not a fielded capability. |
Demonstrate use and care of equipment for tenant use.
14CI 5–23 · exposure 8 · augmentation 38 · importance 3.3/5 · click for rater detail
Demonstrate use and care of equipment for tenant use.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Social service and human service sectors show low overall AI adoption; the populations served (often vulnerable or low-literacy) and the interpersonal nature of the work create structural resistance to automation. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Social services for vulnerable populations (e.g., housing assistance) are a low-digitization, high-touch sector with minimal AI agent deployment for physical task instruction. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist by generating demonstration videos, written guides, or checklists that a human assistant then uses and personalizes, but does not independently transform the core task of live, adaptive demonstration. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could help prepare instructional materials or checklists in advance, but offers little real-time assistance during the actual hands-on demonstration itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Demonstrating use and care of equipment requires physical presence, real-time explanation, and adaptive feedback to individual tenants with varying needs—capabilities current AI cannot reliably provide end-to-end in an in-person setting. AI could assist with creating instructional videos or written guides, but cannot fully replace the interactive human demonstration. |
| Task automatability | claude-sonnet-5 | 1/5 | This is an in-person, hands-on demonstration requiring physical presence, gesture, and real-time observation of tenant comprehension, which current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Direct human contact and trust-building are often essential in social service contexts; tenants may require reassurance, adaptive communication, and hands-on assistance that regulations and best practice strongly favor a human provider to deliver. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No formal licensing is typically required, but the task involves direct human interaction, trust-building, and physical presence that create practical friction against automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The AI cost of producing high-quality demonstration content and interactive guides is non-trivial, while the human wage for a social service assistant is relatively modest, making full replacement economically marginal at best. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute the physical demonstration component, so any AI cost is additive rather than substitutive, making it more expensive than simply having a human do it. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs live, in-person equipment demonstrations to tenants today. While instructional videos and chatbots exist, they do not perform the full task of hands-on demonstration and personalized guidance in real time. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product physically demonstrates equipment use and safety practices to tenants in a residential/social service setting; this remains outside current product capability. |
Consult with supervisor concerning programs for individual families.
6CI 0–11 · exposure 0 · augmentation 38 · importance 3.5/5 · click for rater detail
Consult with supervisor concerning programs for individual families.
6| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Social services organizations operate in highly regulated, human-centric contexts with limited digitization of core workflows; supervisory consultation remains a protected human function with minimal AI adoption pressure. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a sector with historically slow AI adoption for direct case decision-making, though some agencies pilot AI for documentation support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide minor support via summarizing case files or flagging relevant policies for review, but the core consultation—exchanging views, making judgments, and directing action—remains inherently interpersonal and supervisory in nature. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help by summarizing case notes, flagging risk indicators, or preparing briefing materials ahead of the consultation, improving efficiency without replacing the discussion itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires real-time, dynamic consultation between a human assistant and their supervisor about individualized family programs, involving nuanced judgment, contextual understanding, and collaborative decision-making that current AI cannot perform end-to-end at 50% time savings. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a live, judgment-based professional consultation about specific family circumstances requiring nuanced interpersonal discussion and case-specific judgment that AI cannot conduct end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task is deeply embedded in organizational hierarchy and accountability structures; supervisors retain legal and fiduciary responsibility for program oversight and decisions, creating a hard barrier to automation or substitution. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Case oversight typically involves supervisory accountability, confidentiality obligations, and agency/regulatory requirements for professional judgment in family services, creating strong barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Automating this task would require replacing or mediating supervisor-employee consultation, which has minimal direct cost compared to wages but cannot be displaced by inference without eliminating the supervisory function entirely. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | There is no viable AI substitute performing this consultation itself, so cost comparison favors the human process; only ancillary note-taking or summarization tools might reduce prep time cheaply. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs supervisory consultation on individualized family programs; this requires human authority, accountability, and contextual authority vested in the supervisor role that AI systems cannot substitute. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product replaces supervisor-caseworker consultations about individual family cases; this remains a human-to-human interaction embedded in case management workflows. |
Visit individuals in homes or attend group meetings to provide information on agency services, requirements, or procedures.
4CI 0–7 · exposure 0 · augmentation 38 · importance 4.0/5 · click for rater detail
Visit individuals in homes or attend group meetings to provide information on agency services, requirements, or procedures.
4| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Social and human service roles operate in sectors with low automation adoption, heavy dependence on in-person delivery, and regulatory constraints on non-human provision of services. These are labor-intensive, relationship-driven roles where automation velocity remains minimal. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Social services is a historically low-digitization sector with limited AI deployment for direct client-facing physical interactions, though some digital tools are emerging for scheduling/info dissemination. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While AI could assist in preparing information materials, scheduling, or documentation, the core task of visiting and meeting with clients in person offers limited augmentation opportunities, since the human must perform the visit regardless and AI cannot meaningfully enhance the in-person interaction itself. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help assistants prepare materials, summarize agency policies, generate talking points, or draft follow-up communications, improving efficiency around the visit itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires in-person visits to homes and group meetings, which demands physical presence, relationship-building, and contextual responsiveness that current AI systems cannot perform end-to-end. The human contact element is fundamental to the task. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical presence in homes or at meetings, in-person rapport building, and adaptive interpersonal communication that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Significant legal, regulatory, and organizational barriers protect this work: direct client contact is typically required by regulation, liability attaches to the human provider, and the task involves mandatory reporting and assessment responsibilities that require human licensure or authorization in most jurisdictions. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Trust, safety, confidentiality, and often statutory requirements for in-person contact with vulnerable populations create strong barriers to any non-human substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI systems cannot substitute for the physical presence, interpersonal interaction, and trust-building required in home visits and group settings, making comparative cost analysis moot—substitution is not feasible. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical/social component at all, so the relevant cost comparison strongly favors the human worker who must be present. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can conduct home visits or attend in-person group meetings to deliver information and build rapport. This remains entirely within the domain of human service delivery; no production systems perform this task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product conducts home visits or attends group meetings to represent an agency; this remains entirely a human physical and social activity. |
Oversee day-to-day group activities of residents in institution.
3CI 0–5 · exposure 0 · augmentation 25 · importance 4.1/5 · click for rater detail
Oversee day-to-day group activities of residents in institution.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Social service and institutional care sectors are slow to digitize and adopt autonomous systems due to regulatory constraints, the critical importance of human trust and relationship, and the physical nature of work in congregate care settings. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Human/social services and residential care are low-digitization, high-touch sectors with minimal AI agent deployment for direct supervisory care tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide minor assistance through activity scheduling, automated incident logging, or alerting staff to concerning patterns, but the core task of human presence and real-time decision-making cannot be meaningfully augmented by current systems; the human remains fully responsible. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could help with scheduling activities, documentation, or flagging behavioral patterns, but offers limited direct assistance to the core moment-to-moment oversight task. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Day-to-day oversight of resident group activities requires real-time situational awareness, dynamic response to behavioral incidents, and nuanced interpersonal judgment that current AI systems cannot provide end-to-end. While AI might assist with scheduling or documentation, it cannot replace the continuous, adaptive human presence required for resident safety and engagement. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires continuous in-person supervision, real-time judgment, and physical presence to manage group dynamics, safety, and behavioral issues among residents—no AI system can perform this end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Institutional care is heavily regulated and typically requires licensed or credentialed human staff to be physically present and legally responsible for resident welfare and safety. Many jurisdictions mandate human oversight and make automated systems liable for failure, creating hard barriers to substitution. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Institutional care settings typically require staff with specific training/certification, duty-of-care and liability obligations, and residents expect human contact and supervision, creating strong barriers to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The loaded human wage for a social service assistant is modest, and the cost of AI systems capable of embodied oversight (robot, computer vision, 24/7 monitoring infrastructure) plus integration and liability coverage would exceed human labor costs by a large margin. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI product performing this physical, supervisory task, so cost comparison favors humans entirely since AI cannot substitute for the required physical presence and interpersonal oversight. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed products can reliably oversee group activities of residents in institutional settings independently. The task demands embodied presence, immediate crisis response, and contextual understanding of individual residents' needs that exceed current AI capabilities in production environments. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product oversees in-person group activities of residents in institutional settings; this remains firmly in the human domain with no research-stage substitute even attempting full task coverage. |
Transport and accompany clients to shopping areas or to appointments, using automobile.
3CI 0–5 · exposure 0 · augmentation 25 · importance 2.8/5 · click for rater detail
Transport and accompany clients to shopping areas or to appointments, using automobile.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Social service organizations are typically resource-constrained, digitally lagging sectors with strong human-centered missions; automation of client transportation is not a priority in current adoption patterns. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Social services and in-person client transport are low-digitization, physically-grounded sectors with minimal AI adoption for this specific function. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | While GPS routing and scheduling software can assist route planning, the core task of physically driving and accompanying clients offers limited opportunities for meaningful AI augmentation beyond logistics optimization. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can help with route planning, scheduling, or reminders, but offers little assistance to the core physical transport and companionship task itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical presence, real-time navigation in variable environments, and direct human interaction with vulnerable clients. Current AI cannot operate vehicles autonomously in general conditions or provide the in-person accompaniment and support that defines this task. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical driving and personal escort task requiring a human driver and physical presence; current AI cannot perform this end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legal liability for transporting vulnerable clients, regulatory requirements for vehicle operation and passenger safety, duty-of-care obligations, and organizational risk aversion create strong legal and institutional barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Liability, safety regulations, insurance requirements, and the need for a trusted, vetted individual to accompany vulnerable clients create strong barriers to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of autonomous vehicles, insurance, and the inability to eliminate human oversight and accompaniment makes this far more expensive than employing a human assistant to drive and accompany clients. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this task, so cost comparison favors the human by default; any autonomous vehicle solution would still require added infrastructure and oversight costs exceeding current human wages for this narrow task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product today can reliably perform door-to-door transportation with client accompaniment for vulnerable populations. Autonomous vehicles exist only in limited controlled settings and cannot provide the human assistance component. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product transports or physically accompanies clients; autonomous vehicles remain limited, non-standard, and do not include the interpersonal accompaniment aspect. |
Related occupations — Community & Social Service
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.