Child, Family, and School Social Workers

21-1021.00
Median wage $59,550/yr392,550 employed (US)Rank #818 of 923 scored · top 89% by substitution

Provide social services and assistance to improve the social and psychological functioning of children and their families and to maximize the family well-being and the academic functioning of children. May assist parents, arrange adoptions, and find foster homes for abandoned or abused children. In schools, they address such problems as teenage pregnancy, misbehavior, and truancy. May also advise teachers.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution14
Exposure14
Augmentation49

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

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

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

Tasks on the substitution scale

21 rated tasks, binned by substitution score.

Position among all scored occupations

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

Tasks with substitution ≥ 70

0%

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

Why this score

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

Task automatabilityw 35%15

panel mean rating 1.6/5 → substitution pressure 15/100

Technical feasibility todayw 20%12

panel mean rating 1.5/5 → substitution pressure 12/100

Cost vs. human wagew 15%17

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

Adoption barriersw 20%inverted — strong barriers lower the score13

panel mean rating 4.5/5 (barrier strength) → substitution pressure 13/100

Sector adoption velocityw 10%12

panel mean rating 1.5/5 → substitution pressure 12/100

Task breakdown (21 tasks)

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

Maintain case history records and prepare reports.

57

CI 55–60 · exposure 66 · augmentation 75 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Social work agencies lag in digitization and AI adoption; most operate on legacy systems, and resource constraints limit investment in new tools; adoption remains primarily in larger urban agencies and is slow relative to information-sector benchmarks.
Sector adoption velocityclaude-sonnet-52/5Social services is a historically low-digitization, under-resourced sector with slow technology adoption compared to finance or professional services, though some agencies pilot AI documentation tools.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-assisted drafting of case notes and automated report generation can substantially accelerate documentation while a licensed social worker reviews and signs, freeing time for direct client contact and clinical work.
Augmentation potentialclaude-sonnet-54/5AI can substantially speed up drafting, organizing, and summarizing case histories while the social worker remains responsible for accuracy, judgment, and final sign-off.
Task automatabilityclaude-haiku-4-5-202510014/5AI systems can effectively draft case notes, organize historical records, and auto-populate structured report templates with information from interviews and documents. End-to-end automation with ≥50% time savings is achievable, though mandatory human review and signature requirements prevent truly autonomous completion.
Task automatabilityclaude-sonnet-54/5Drafting case notes, summarizing sessions, and generating structured reports from dictated or written input is well within current LLM capability, especially with templates and structured intake data.
Adoption barriersclaude-haiku-4-5-202510014/5Legal and regulatory barriers are significant: licensed social workers must typically sign off on case records and reports; liability for documentation errors, duty-of-care requirements, and institutional compliance with child protection laws create strong barriers to full automation.
Adoption barriersclaude-sonnet-53/5Records often require professional certification, accurate legal/clinical language, and confidentiality compliance (e.g., HIPAA, child welfare regulations), creating moderate friction even though the drafting itself isn't legally restricted to a licensed person.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI inference and integration costs for document processing and report generation are substantially lower than the loaded cost of human social workers' clerical time; savings are particularly pronounced for high-volume templated documentation.
Cost vs. human wageclaude-sonnet-54/5AI drafting of routine case notes and reports is far cheaper per document than a social worker's time, though oversight and correction still require paid professional review.
Technical feasibility todayclaude-haiku-4-5-202510013/5Document drafting and report generation tools exist in production (legal tech, healthcare EHR systems), but social work-specific solutions with demonstrated reliability remain limited; deployed products often require substantial human editing and face accuracy concerns in sensitive contexts.
Technical feasibility todayclaude-sonnet-53/5Some case management software includes AI-assisted note generation and summarization, but adoption in social work agencies is uneven and human review/editing remains standard due to confidentiality and accuracy concerns.

Collect supplementary information needed to assist client, such as employment records, medical records, or school reports.

52

CI 25–80 · exposure 58 · augmentation 88 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Social services and schools are mid-digitization sectors with uneven adoption of advanced document automation. While some large agencies use RPA and AI-assisted record retrieval, smaller and rural agencies lag; overall adoption is pilot-to-early-production rather than widespread.
Sector adoption velocityclaude-sonnet-52/5Social work and social services are among the slower-adopting sectors for AI, with limited digitization of inter-agency data exchange and heavy reliance on manual, relationship-based information gathering.
Augmentation potentialclaude-haiku-4-5-202510015/5AI document retrieval and summarization tools significantly augment social workers by rapidly surfacing relevant records, flagging inconsistencies, and organizing information for faster case assessment, freeing workers to focus on client interaction and decision-making.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist by auto-drafting record request letters, tracking outstanding documents, extracting and summarizing key details from received records, and flagging missing information, improving caseworker efficiency substantially while they remain in charge of legal compliance and follow-up.
Task automatabilityclaude-haiku-4-5-202510015/5Collecting and organizing supplementary information from standard sources (employment records, medical records, school reports) is largely a matter of retrieving, parsing, and aggregating documents—tasks AI excels at with existing document processing, OCR, and database query tools. Current systems can handle this end-to-end with significant time savings.
Task automatabilityclaude-sonnet-52/5Gathering records involves outreach, coordination with multiple institutions, and follow-up communication that AI cannot fully execute autonomously, though it can help draft requests and organize received documents. It falls short of the 50% end-to-end time-saving bar because much of the effort is inter-organizational coordination and consent-based access requiring human judgment.
Adoption barriersclaude-haiku-4-5-202510013/5HIPAA, FERPA, and state data protection laws govern access to medical and educational records, requiring careful authorization and privacy controls. However, these are compliance/oversight barriers rather than legal prohibitions on automation itself; a social worker can legally delegate retrieval to a compliant system.
Adoption barriersclaude-sonnet-54/5Records covered by HIPAA, FERPA, and confidentiality laws require signed consent and often direct human verification, creating substantial legal and privacy barriers to full automation of information collection.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI-driven document collection and aggregation costs a fraction of manual record gathering by human staff, including eliminated phone calls, faxes, and administrative overhead. Labor displacement is substantial relative to loaded wages.
Cost vs. human wageclaude-sonnet-52/5While AI can cheaply summarize or organize records once obtained, the actual collection process still requires human phone calls, signed releases, and institutional coordination, so the all-in cost savings versus a human worker are modest.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed products (document management systems, AI-powered record retrieval, and RPA solutions) reliably extract and organize structured information from employment, medical, and educational records in production healthcare and social services settings. Minor friction remains around legacy format interoperability and authorization verification.
Technical feasibility todayclaude-sonnet-52/5No deployed product autonomously collects employment, medical, and school records for a caseworker; existing tools are limited to document management or e-fax/EHR integrations that still require human-initiated requests and verification.

Determine clients' eligibility for financial assistance.

36

CI 25–48 · exposure 38 · augmentation 63 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Social service agencies are typically low-digitization, under-resourced organizations with legacy systems. Adoption of AI automation in this space is minimal and slow due to funding constraints, risk-aversion around liability, and limited vendor focus on social work workflows.
Sector adoption velocityclaude-sonnet-52/5Public social services and school-based social work are traditionally slow-adopting sectors with limited IT budgets and cautious rollout of automated decision tools, especially for vulnerable populations.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by extracting and summarizing client documentation, flagging potential eligibility pathways, and drafting initial eligibility assessments for human review. Such tools would raise worker productivity on routine cases while leaving final determination to the social worker.
Augmentation potentialclaude-sonnet-54/5AI can efficiently pre-screen applications, flag missing documentation, and summarize case files, meaningfully speeding up the social worker's eligibility review process while they retain final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5Eligibility determination involves complex rule interpretation, evidence verification, and discretionary judgment about client circumstances. While basic rule-matching could be automated, verifying documentation, handling edge cases, and interpreting ambiguous situations require human oversight that prevents the 50% time-saving threshold at equal quality.
Task automatabilityclaude-sonnet-53/5Eligibility determination against defined criteria (income thresholds, documentation checks) can be substantially automated via rules engines and AI document processing, but nuanced case circumstances and exceptions still require human judgment.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: legal liability for incorrect eligibility determinations, regulatory requirements for documented decision-making and appeals processes, and client-contact and judgment requirements embedded in social work licensing standards and agency protocols.
Adoption barriersclaude-sonnet-53/5No licensing requirement mandates a human specifically for eligibility determination, but agency policy, due process rights, appeals mechanisms, and accountability for wrongful denials create real friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Integration costs, compliance oversight, and the need for human verification of AI outputs make automation expensive relative to having a social worker perform or spot-check eligibility determination, particularly in resource-constrained social service agencies.
Cost vs. human wageclaude-sonnet-53/5Automated eligibility checks are cheaper per case for straightforward criteria, but the need for human oversight, appeals handling, and case-specific judgment keeps blended costs closer to parity rather than order-of-magnitude savings.
Technical feasibility todayclaude-haiku-4-5-202510012/5No mature product reliably performs end-to-end eligibility determination for financial assistance in social work settings. Narrow tools exist for specific programs or institutions, but they typically require significant customization, human review, and cannot handle the full range of client situations and documentation types.
Technical feasibility todayclaude-sonnet-53/5Eligibility screening software and AI-assisted document verification are deployed in benefits administration, but full end-to-end determination in social work contexts still typically involves human review due to complex family circumstances.

Conduct social research.

32

CI 25–39 · exposure 25 · augmentation 63 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Social work organizations, schools, and family service agencies operate in traditionally lower-digitization sectors with risk-averse institutional cultures; adoption of AI-driven research is nascent and limited mainly to data analysis tools rather than end-to-end automation.
Sector adoption velocityclaude-sonnet-52/5Social services and nonprofit sectors historically lag in AI adoption due to funding constraints, data sensitivity, and limited digital infrastructure, resulting in slow uptake for research support tools.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist social researchers by automating literature searches, coding qualitative data, and performing statistical analysis, raising human productivity on specific subtasks while the researcher retains control over design, ethics, and interpretation.
Augmentation potentialclaude-sonnet-54/5AI significantly aids literature synthesis, drafting research questions, summarizing case data, and identifying patterns in qualitative data, meaningfully boosting a social worker's research productivity while they retain oversight.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist with literature review, data collection, and analysis, conducting social research requires interpreting nuanced human contexts, designing studies ethically, and making judgments about vulnerable populations that are not yet automatable end-to-end at 50% time savings with equal quality.
Task automatabilityclaude-sonnet-52/5AI can accelerate literature review, data synthesis, and drafting of research summaries, but designing valid social research, ensuring ethical sampling, and interpreting community-specific context still require substantial human judgment and fieldwork.4 Only partial subtasks meet the 50% time-savings bar.
Adoption barriersclaude-haiku-4-5-202510014/5Social research involving children and families is subject to IRB oversight, informed consent requirements, and ethical compliance mandates that legally require human researcher judgment and accountability; liability for research design flaws and harm to participants creates strong barriers to full automation.
Adoption barriersclaude-sonnet-52/5No licensing requirement specifically restricts AI from supporting research tasks, though professional standards and IRB/ethical review processes for research involving vulnerable populations create moderate friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI tools for data analysis and literature review are relatively inexpensive, but the human expertise required for research design, ethical decision-making, and interpretation of findings means the all-in cost remains comparable to or higher than a human researcher.
Cost vs. human wageclaude-sonnet-53/5AI can cheaply handle background research and data summarization, lowering costs for parts of the task, but human-led design, interviews, and validation still dominate the cost structure, keeping overall savings moderate.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some AI tools exist for literature mining and statistical analysis, but no deployed product reliably performs the full research design, ethical oversight, stakeholder engagement, and contextual interpretation required for social research in child/family settings.
Technical feasibility todayclaude-sonnet-52/5Tools like literature-review assistants and data analysis copilots exist but are not deployed as end-to-end social research systems for social work practice; usage is ad hoc rather than production-grade for this specific task.

Refer clients to community resources for services, such as job placement, debt counseling, legal aid, housing, medical treatment, or financial assistance, and provide concrete information, such as where to go and how to apply.

29

CI 25–34 · exposure 25 · augmentation 75 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Social services remain low-digitization, relationship-heavy sectors with fragmented funding and organizational structures. Adoption of AI agents in referral workflows is minimal; most jurisdictions still rely on manual resource navigation and human social worker judgment.
Sector adoption velocityclaude-sonnet-52/5Social services is a historically low-digitization, resource-constrained sector with slow AI adoption, mostly limited to pilot chatbots or resource databases rather than integrated production workflows.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can usefully assist social workers by surfacing relevant resources, filtering eligibility criteria, maintaining searchable directories, and tracking application steps—significantly reducing information-gathering burden. However, the human social worker must remain in the loop for judgment, relationship-building, and suitability assessment.
Augmentation potentialclaude-sonnet-54/5AI tools can significantly speed up finding relevant community resources, drafting referral information, and organizing application steps, meaningfully boosting caseworker efficiency while they remain the decision-maker.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can retrieve and organize information about community resources and eligibility criteria, the task requires nuanced judgment about individual client needs, circumstances, and suitability of referrals—plus navigation of complex, frequently changing resource availability. Current systems can assist with information compilation but cannot reliably make context-sensitive referral decisions or handle the interpersonal calibration needed.
Task automatabilityclaude-sonnet-52/5AI can help identify and compile relevant resources and generate application instructions, but the referral requires client-specific judgment, relationship trust, and follow-up that current systems cannot fully replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Social work referrals carry liability and duty-of-care requirements; inappropriate referrals can harm vulnerable clients. Professional licensing and legal liability for bad outcomes create strong barriers, and clients often require human relationship and advocacy that cannot be automated. Regulatory frameworks and ethical standards expect licensed professionals to exercise discretion.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement to share resource information, but professional standards, liability for bad referrals, and client trust/vulnerability create meaningful friction against pure automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI tools for resource databases and basic matching are cheap, but the oversight and verification required to ensure safe, appropriate referrals—plus the need for human social workers to validate and contextualize recommendations—means all-in cost remains competitive with or higher than human-alone performance.
Cost vs. human wageclaude-sonnet-53/5AI-assisted resource lookup is cheap compared to a caseworker's time, but the overall task still requires human verification, personalization, and follow-through, limiting overall savings.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs holistic referral triage and resource-matching for social services at production scale. Some databases and directory tools exist, but they lack the contextual reasoning, relationship history, and individualized matching that this task demands. Systems that attempt this remain in pilot phases with significant error rates.
Technical feasibility todayclaude-sonnet-52/5Chatbots and resource databases exist to surface referral information, but no deployed product reliably handles the full referral process including judgment calls about client needs and appropriateness in production social work settings.

Develop and review service plans in consultation with clients and perform follow-ups assessing the quantity and quality of services provided.

25

CI 25–25 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Social services sectors show slower AI adoption than information or finance sectors; organizations remain cautious about automating tasks affecting vulnerable populations. Adoption is mostly in back-office automation (scheduling, documentation) rather than core service planning and assessment tasks.
Sector adoption velocityclaude-sonnet-52/5Social services is a historically slow-adopting sector with limited digitization and cautious AI integration, especially in sensitive child and family contexts.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by drafting plan templates, flagging service gaps in existing data, and generating assessment summaries from case notes. However, the human social worker must synthesize client input, make clinical judgments, and own the relationship, so augmentation is helpful but not transformative of the core task.
Augmentation potentialclaude-sonnet-53/5AI can help draft service plan documents, organize case notes, and flag follow-up items, providing moderate productivity gains while the social worker remains central to client interaction and judgment.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can draft service plans from templates and generate assessment reports, the task fundamentally requires collaborative consultation with clients, judgment about appropriateness of services, and relationship-based follow-up. These cannot be fully automated; human oversight and client engagement remain essential, limiting time savings below the 50% threshold.
Task automatabilityclaude-sonnet-52/5Developing service plans requires nuanced client interaction, judgment about family circumstances, and relationship-based trust that current AI cannot replicate end-to-end; AI can only assist with drafting or documentation portions.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: social work licensing requirements mandate that qualified professionals conduct assessments and develop plans, liability exposure is high if services are inadequate, and regulatory frameworks require documented human judgment in child/family welfare contexts. Client trust and mandated human-client contact further protect this role.
Adoption barriersclaude-sonnet-54/5This task typically requires a licensed social worker to exercise professional judgment, maintain confidentiality, and be accountable for child welfare decisions, creating strong regulatory and liability barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Implementation of specialized case management AI, compliance infrastructure, and human oversight adds meaningful cost relative to traditional record-keeping. The human social worker remains the primary cost driver and cannot be removed, so the cost ratio remains unfavorable for AI substitution.
Cost vs. human wageclaude-sonnet-52/5While AI drafting tools reduce documentation time cheaply, the core consultation and assessment work still requires a paid licensed social worker, keeping overall cost close to human-driven levels.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI-powered documentation and case management tools exist but do not perform independent consultation or quality assessment reliably. Current products assist with record-keeping and generating templates, but real-world deployment still requires social workers to lead all client interactions and make clinical judgments about service adequacy.
Technical feasibility todayclaude-sonnet-52/5Some case management software includes templated plan generation and follow-up scheduling, but no deployed product independently develops or reviews service plans with clients reliably in production.

Arrange for medical, psychiatric, and other tests that may disclose causes of difficulties and indicate remedial measures.

25

CI 25–25 · exposure 25 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Child welfare and school social work remain heavily manual, relationship-driven fields with limited digitization. Adoption of AI agents in production is rare; most adoption is confined to documentation and record-keeping rather than care-coordination or test arrangement.
Sector adoption velocityclaude-sonnet-52/5Social services and school-based settings are generally slow AI adopters due to funding constraints, sensitive data concerns, and reliance on human judgment in child welfare contexts.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by suggesting relevant tests based on case notes, flagging standard assessments, and summarizing clinical pathways, helping social workers make more informed recommendations. However, the human social worker must retain judgment and coordinate execution.
Augmentation potentialclaude-sonnet-53/5AI can assist by helping identify relevant tests/specialists, generating referral letters, and organizing case documentation, meaningfully supporting but not replacing the social worker's coordination role.
Task automatabilityclaude-haiku-4-5-202510012/5AI can help identify which tests to recommend based on symptom analysis, but arranging tests requires coordination with healthcare providers, insurance authorization, and scheduling—tasks heavily dependent on human judgment, institutional access, and real-world logistics that current AI systems cannot execute end-to-end.
Task automatabilityclaude-sonnet-52/5AI can help identify appropriate tests/referrals and draft paperwork, but arranging tests requires coordination with providers, judgment calls, and human relational trust that current systems cannot fully replace end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Social workers must be licensed professionals in most jurisdictions, and test arrangement often requires their documented recommendation and follow-up. Healthcare providers typically require a qualified professional's clinical judgment, and liability for missed diagnoses creates strong barriers to full automation.
Adoption barriersclaude-sonnet-54/5Social workers often need licensure and legal authority to make referrals and coordinate care for minors, with liability and confidentiality (HIPAA/FERPA) constraints creating strong barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI inference for test recommendation is cheap, but the coordination, insurance navigation, provider communication, and scheduling still require human social worker time. The total cost of automation would likely exceed the cost of a social worker performing the task, given the human touchpoints necessary.
Cost vs. human wageclaude-sonnet-52/5AI could cut some administrative time (referral drafting, scheduling logistics) but the core judgment and coordination work still requires a paid professional, limiting cost savings.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can draft test recommendations or summarize clinical guidelines, no deployed product reliably arranges actual medical tests across diverse healthcare systems without significant human intervention. AI tools exist for documentation and triage but lack the institutional integration and decision authority required for this task.
Technical feasibility todayclaude-sonnet-52/5No deployed product autonomously arranges medical/psychiatric evaluations for at-risk children; scheduling and referral tools exist but require heavy human oversight and case-specific judgment.

Evaluate personal characteristics and home conditions of foster home or adoption applicants.

10

CI 0–20 · exposure 13 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Child welfare agencies are laggards in AI adoption due to high liability, regulatory constraints, and the mission-critical nature of decisions. Pilot projects exist, but production deployment of autonomous evaluation systems remains rare and cautious.
Sector adoption velocityclaude-sonnet-51/5Social work and child welfare agencies are slow, resource-constrained, and heavily regulated, with minimal AI deployment for in-person assessments.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by extracting and organizing information from application documents, flagging missing data, and preparing summary reports that social workers review and validate. This augmentation improves efficiency without removing human judgment from the evaluation.
Augmentation potentialclaude-sonnet-53/5AI can help with drafting reports, organizing case notes, and flagging risk factors from structured data, but cannot replace the core judgment and observation of the visit.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with document review and initial screening of application materials, but the task requires nuanced judgment of personal characteristics, family dynamics, and home safety—areas where current AI has high error rates and cannot reliably replicate the 50% time-saving threshold at equal quality. Human-led evaluation with AI-aided data summarization is feasible, but end-to-end automation remains infeasible.
Task automatabilityclaude-sonnet-51/5This requires in-person home visits, interpersonal judgment, observation of family dynamics, and safety assessments that current AI cannot perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5Strong legal and regulatory barriers exist: licensure laws require qualified social workers to conduct assessments, liability falls on credentialed professionals, and child welfare statutes mandate human judgment and accountability. Automation of the core decision-making is heavily restricted by law and professional regulation.
Adoption barriersclaude-sonnet-55/5Licensed social workers are legally required to conduct and certify these evaluations for child welfare and court proceedings, creating hard regulatory and liability barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5Integration costs (training, data infrastructure, oversight by licensed social workers) are substantial, and the AI system would require significant human review and correction, keeping all-in costs close to or above the cost of human social workers performing the task directly.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the physical home visit and interpersonal assessment, so there is no meaningful AI cost comparison—human labor is required.
Technical feasibility todayclaude-haiku-4-5-202510012/5No production systems reliably perform comprehensive foster/adoption suitability assessments independently. While AI can flag obvious documentation gaps or assist with risk scoring, real organizations still require trained social workers to conduct in-person visits, make welfare judgments, and take legal responsibility for placement decisions.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product conducts home evaluations or assesses applicant suitability for foster/adoption placement; this remains a human field-based task.

Counsel individuals, groups, families, or communities regarding issues including mental health, poverty, unemployment, substance abuse, physical abuse, rehabilitation, social adjustment, child care, or medical care.

9

CI 4–15 · exposure 8 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption in social work remains slow. Most agencies rely on licensed staff and are cautious about automation in direct client care due to liability, ethical concerns, and regulatory constraints. Pilot projects exist but production replacement is minimal.
Sector adoption velocityclaude-sonnet-51/5Social work and child welfare services are a low-digitization, high human-contact sector with minimal AI agent deployment in direct counseling roles.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist social workers by organizing case notes, flagging risk factors, suggesting evidence-based resources, and drafting documentation, raising administrative efficiency. However, augmentation in the core counseling interaction itself is limited because the human must assess, decide, and respond to clients directly.
Augmentation potentialclaude-sonnet-53/5AI can assist with case documentation, resource matching, risk-flagging, and preparatory research, meaningfully supporting but not replacing the counseling relationship itself.
Task automatabilityclaude-haiku-4-5-202510012/5Counseling requires nuanced emotional understanding, rapport-building, and real-time adaptation to client needs that current AI cannot reliably achieve. While AI can provide information and draft guidance, it cannot replicate the therapeutic relationship, clinical judgment, and ethical accountability essential to effective counseling.
Task automatabilityclaude-sonnet-51/5Counseling on high-stakes issues like abuse, mental health, and substance abuse requires nuanced human judgment, trust-building, and legal/ethical accountability that current AI cannot replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Counseling is heavily regulated: social workers must hold state licenses, meet ethical standards, maintain confidentiality under law, and assume liability for outcomes. Most jurisdictions legally require a human professional to conduct assessments, treatment planning, and crisis intervention.
Adoption barriersclaude-sonnet-55/5This task typically requires licensed social workers, mandated reporting obligations, and legal accountability for child welfare decisions, creating hard regulatory and liability barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI tools (chatbots, screening systems) cost less per interaction than human counselors, but integration, liability oversight, and regulatory compliance add overhead. For equivalent therapeutic outcome and accountability, the all-in cost remains higher than deployed AI saves.
Cost vs. human wageclaude-sonnet-52/5While AI inference is cheap, the liability, oversight, and quality-assurance costs required to safely deploy AI in place of a licensed social worker make the effective cost comparable or higher than human labor.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs independent counseling at clinical standards. Chatbots exist for crisis support and psychoeducation, but they lack the ability to conduct comprehensive assessments, diagnose, provide evidence-based treatment, or navigate complex trauma—all core to this task.
Technical feasibility todayclaude-sonnet-51/5No deployed products perform independent counseling for child welfare or family crisis situations in production; chatbot mental health tools remain adjunct and narrowly scoped, not substitutes for professional social work counseling.

Interview clients individually, in families, or in groups, assessing their situations, capabilities, and problems to determine what services are required to meet their needs.

6

CI 0–13 · exposure 8 · augmentation 38 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Child welfare and social services sectors are heavily regulated, rely on in-person contact, and move slowly on automation due to liability concerns and the critical nature of decisions. Adoption of AI for core assessment functions remains minimal.
Sector adoption velocityclaude-sonnet-51/5Social services is a historically low-digitization, human-contact-intensive sector with slow AI adoption for core casework, especially for direct client assessment interviews.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist with note-taking or post-interview documentation, but the core task—interactive assessment through dialogue with vulnerable clients—resists meaningful augmentation since the human must remain actively present and in control of clinical judgment throughout.
Augmentation potentialclaude-sonnet-53/5AI can help with note-taking, summarizing case histories, flagging risk indicators, or drafting assessment reports, offering moderate productivity support while the social worker still leads the interview.
Task automatabilityclaude-haiku-4-5-202510011/5Assessing client situations, capabilities, and problems requires nuanced human judgment, empathy, and ability to build trust with vulnerable individuals and families. Current AI systems cannot reliably conduct intake interviews or clinical assessments that meet professional standards for social work practice.
Task automatabilityclaude-sonnet-52/5Conducting sensitive clinical interviews with vulnerable clients requires real-time human rapport, nonverbal cue reading, and trust-building that current AI cannot replicate end-to-end; some transcription/note support exists but the core interviewing and clinical judgment remain human.rated tasks.
Adoption barriersclaude-haiku-4-5-202510015/5Social work assessments are legally and ethically bound to licensed practitioners; many jurisdictions require licensed social workers to conduct initial assessments and have professional liability obligations. Client consent, duty of care, and mandatory reporting laws create hard regulatory barriers to full automation.
Adoption barriersclaude-sonnet-55/5Child welfare and family assessments typically require a licensed social worker's judgment and are governed by mandatory reporting laws, confidentiality regulations, and liability concerns, making human sign-off legally and ethically required.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of AI systems capable of attempting such interviews, plus required human oversight and reassessment of incorrect judgments, would exceed the cost of direct human interviewing by trained social workers.
Cost vs. human wageclaude-sonnet-52/5Even where AI-assisted intake tools exist, they still require licensed staff to conduct and validate the actual assessment, so cost savings are modest rather than order-of-magnitude given liability and oversight needs.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed social work product reliably performs independent client assessment interviews at scale. While chatbots exist, they lack the clinical judgment, ability to detect non-verbal cues, and professional accountability required for real intake assessments in child welfare or family services.
Technical feasibility todayclaude-sonnet-51/5No deployed product conducts autonomous clinical/social work assessment interviews with clients in production; existing chatbots for intake screening are narrow and not trusted for high-stakes family/child welfare assessments.

Consult with parents, teachers, and other school personnel to determine causes of problems, such as truancy and misbehavior, and to implement solutions.

4

CI 0–7 · exposure 0 · augmentation 38 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5School social work is a regulated profession with human licensure requirements and limited digitization of core consulting and problem-solving work. Adoption velocity of AI in this role is negligible; the sector prioritizes face-to-face relationship and accountability over automation.
Sector adoption velocityclaude-sonnet-52/5School social work and K-12 education settings show slow, uneven AI adoption for core casework functions, with mostly administrative/documentation tools being piloted rather than deep integration into consultation work.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist modestly—drafting summary notes, flagging attendance patterns, or organizing information from multiple sources—but the interpretive and relational core of consulting with stakeholders and diagnosing root causes remains firmly human-dependent. Assistive gains are limited to peripheral documentation tasks.
Augmentation potentialclaude-sonnet-53/5AI can help by summarizing case notes, tracking attendance/behavior patterns, drafting communication, and suggesting intervention frameworks, but the core consultative and diagnostic work still depends on human relational skill.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires understanding complex interpersonal dynamics, reading social context, and building trust across multiple stakeholders—capabilities where current AI falls far short. The task involves judgment calls about root causes (emotional, family, developmental, systemic) that demand lived human understanding and the ability to establish credibility with parents and educators.
Task automatabilityclaude-sonnet-51/5This requires live, trust-based, multi-party interpersonal consultation and situational judgment involving vulnerable children; AI cannot conduct these relational conversations or make on-the-spot contextual decisions today.
Adoption barriersclaude-haiku-4-5-202510015/5Strong legal and regulatory barriers protect this task: school social workers typically hold state licensure or certification, carry liability for recommendations affecting minors, and must document interactions in compliance with FERPA and education law. Parents and schools expect a licensed human professional; no substitute bypasses these requirements.
Adoption barriersclaude-sonnet-54/5Professional licensure, confidentiality obligations, child welfare liability, and the necessity of trusted human judgment in sensitive family/school situations create strong barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Current AI systems cannot perform this task end-to-end; the relevant deployed technologies (chatbots, scheduling tools) address only fragments. The human social worker's loaded cost remains far lower than the total cost of building, integrating, and overseeing AI systems that would need to handle stakeholder communication, liability, and intervention design.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this consultation function, so cost comparison favors the human by default; any AI role is confined to support tools, not replacement.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs multi-stakeholder diagnosis and solution implementation for behavioral/attendance problems in schools. While AI can assist with data gathering or documentation, the core task of consulting with diverse parties, inferring causation, and facilitating agreement on solutions remains outside production AI capability.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs live consultations with parents, teachers, and school staff to diagnose behavioral causes and implement solutions; this remains firmly a human relational activity.

Provide, find, or arrange for support services, such as child care, homemaker service, prenatal care, substance abuse treatment, job training, counseling, or parenting classes to prevent more serious problems from developing.

4

CI 0–7 · exposure 0 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Social work remains a highly human-centered, relationship-dependent profession with limited automation adoption even for administrative tasks. Organizations employing social workers are often under-resourced nonprofits or public agencies with slow technology adoption, and the sector has not moved toward agent-based automation in production.
Sector adoption velocityclaude-sonnet-52/5Social services agencies are generally slow to adopt AI due to funding constraints, regulatory sensitivity, and the relational nature of the work, with pilots rare and production deployment minimal.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist social workers by organizing and summarizing available support services, drafting referral documentation, or pulling client history, raising efficiency on administrative and research components of the role while the social worker retains assessment and relationship responsibilities.
Augmentation potentialclaude-sonnet-53/5AI can help by searching for available resources, generating referral lists, drafting case notes, or summarizing eligibility criteria, improving efficiency while the social worker retains judgment and client interaction.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires complex judgment about individual circumstances, relationship-building, understanding of community resources, and discretionary decisions about what services to arrange. While AI could assist with information retrieval about available services, the core responsibility—determining appropriate interventions, evaluating client readiness, and arranging personalized support—requires human social work expertise and cannot be automated end-to-end at the 50% time-saving threshold.
Task automatabilityclaude-sonnet-51/5This task requires in-person assessment, relationship-building, judgment about family circumstances, and coordination with multiple human agencies, none of which current AI can perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Strong legal and regulatory barriers protect this work: social workers must be licensed professionals (LSW/LCSW), client confidentiality is governed by strict privacy laws (FERPA, HIPAA), and liability for case decisions rests with the licensed worker. Clients—especially vulnerable families and children—have strong preferences for human relationships, and many interventions require in-person assessment and trust-building that cannot be automated.
Adoption barriersclaude-sonnet-54/5Social work involving child welfare and family services is typically governed by licensing, mandated reporting duties, and legal liability, requiring a credentialed professional to make judgments and sign off on service plans.
Cost vs. human wageclaude-haiku-4-5-202510011/5Even accounting for AI-assisted research into resources, the human social worker wage for comprehensive case management, assessment, and coordination far exceeds the cost of AI tools that could only handle narrow subtasks like database queries or document drafting.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the human labor of assessing needs, building trust, and coordinating services, so there is no viable AI-only cost comparison; a human must still perform the core work.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product reliably performs the full task of assessing client needs, selecting appropriate interventions, and coordinating multi-faceted support services in production settings. While AI chatbots exist for mental health information and resource databases are searchable, real social work requires human judgment, empathy, and accountability that current systems cannot demonstrate reliably.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product independently arranges wraparound social services for at-risk families; existing tools are limited to referral databases or scheduling aids, not case coordination.

Serve on policy-making committees, assist in community development, and assist client groups by lobbying for solutions to problems.

4

CI 0–7 · exposure 0 · augmentation 38 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI for this task is negligible because the task's core function—legitimate human representation in governance and policy spaces—cannot be delegated to machines. Sectors performing this work have not and cannot meaningfully substitute AI for the human participation requirement.
Sector adoption velocityclaude-sonnet-52/5Social work and community organizing sectors show slow AI adoption for advocacy and policy work, with AI mainly used for case documentation rather than representation or lobbying.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can provide limited augmentation through policy research synthesis, data analysis supporting advocacy positions, and draft communication materials. However, the human-centric nature of committee participation and lobbying limits how much AI assistance can meaningfully enhance the core task.
Augmentation potentialclaude-sonnet-53/5AI can help draft policy briefs, summarize community data, research precedents, and prepare talking points, meaningfully supporting the human advocate's preparation and communication.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires authentic human representation of client interests, persuasion of elected officials and community stakeholders, and real-time relationship-building in policy environments. AI cannot credibly serve on policy-making committees or legitimately lobby on behalf of client groups, as these activities require legal standing, accountability, and genuine advocacy presence.
Task automatabilityclaude-sonnet-51/5This task requires representing human interests, building trust, political negotiation, and in-person community engagement that AI cannot perform end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Strong legal and regulatory barriers exist: lobbying activities are regulated, policy committee membership requires designated human representatives with accountability, and client advocacy requires authenticated human agency. These are hard barriers to substitution.
Adoption barriersclaude-sonnet-54/5Serving on committees and lobbying often requires formal representation, credentialing, and accountable human presence; organizations and communities expect a real, licensed advocate rather than an AI system.
Cost vs. human wageclaude-haiku-4-5-202510011/5The core value of this task—human advocacy presence and credibility in policy spaces—cannot be replaced by AI inference. Oversight and human participation are non-negotiable, making the effective cost-per-outcome prohibitively expensive compared to direct human involvement.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this role, so the comparison is moot—human social workers remain the only viable option, making AI effectively more costly since it cannot deliver the output at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system can authentically participate as a voting member of policy committees or conduct lobbying activities that require human legal authority and accountability. While AI can draft policy briefs, these are support tools, not autonomous task completion.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product serves on policy committees, engages in community organizing, or lobbies on behalf of client groups; this remains a human relational and political function.

Counsel students whose behavior, school progress, or mental or physical impairment indicate a need for assistance, diagnosing students' problems and arranging for needed services.

3

CI 0–5 · exposure 5 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5School social work and clinical counseling occur in highly regulated, relationship-dependent environments with strong professional licensing requirements and institutional risk aversion; adoption of AI for core counseling tasks remains minimal in practice.
Sector adoption velocityclaude-sonnet-51/5School social work is a low-digitization, high-touch human service sector with minimal AI agent deployment in direct student counseling contexts.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist with screening questionnaires, documentation, resource recommendations, or psychoeducational materials to support a human social worker's workflow, but augmentation is limited by the primacy of human judgment and therapeutic presence in this task.
Augmentation potentialclaude-sonnet-53/5AI can help with documentation, resource lookup, drafting referral letters, and flagging risk indicators from case notes, but the core counseling and diagnostic judgment remains human-led.
Task automatabilityclaude-haiku-4-5-202510011/5Counseling students requires nuanced emotional intelligence, relationship-building, clinical judgment, and diagnosis of complex psychosocial problems that current AI cannot reliably perform end-to-end. While AI can assist with screening or information gathering, the core therapeutic and diagnostic work demands human expertise and presence.
Task automatabilityclaude-sonnet-51/5Direct counseling of children involving diagnosis and personalized service arrangement requires clinical judgment, relational trust, and ethical accountability that current AI cannot replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Counseling and diagnosis of minors is legally and ethically bound to licensed social workers or mental health professionals; liability, informed consent, mandated reporting, and child protection laws create hard barriers to substitution by AI systems.
Adoption barriersclaude-sonnet-55/5This work typically requires licensure (LCSW or similar), mandated reporting duties, and legal/ethical responsibility for diagnosis and referral that only a credentialed human can fulfill.
Cost vs. human wageclaude-haiku-4-5-202510011/5Deploying AI for counseling would still require clinical human oversight, training, integration, and liability management; total cost would exceed the alternative of direct human social worker intervention, and diagnostic errors carry high correction costs.
Cost vs. human wageclaude-sonnet-51/5Given liability, need for human oversight, and lack of viable autonomous AI substitute, the effective all-in cost of an AI system would exceed or match the human social worker's cost for equivalent trusted output.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs psychological counseling, diagnosis, or case management at production scale. AI tools exist for screening and psychoeducational content, but clinical diagnosis and therapeutic intervention remain within human professional domains with no mature alternative systems.
Technical feasibility todayclaude-sonnet-51/5No deployed products autonomously diagnose student mental/behavioral issues and arrange services; existing AI mental-health tools are adjunctive and not used this way with minors in schools.

Supervise other social workers.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Social work agencies operate in regulated, hierarchical environments where supervisory authority rests on human accountability and professional judgment. Adoption of AI to replace supervision is negligible because the legal and ethical framework forbids it.
Sector adoption velocityclaude-sonnet-52/5Social services is a sector with generally low AI adoption for core clinical/managerial functions, with pilots limited to administrative support rather than supervisory replacement.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with case documentation, performance data aggregation, or scheduling, but supervisory decision-making—coaching, disciplinary action, professional development planning—remains fundamentally human and cannot be augmented by AI in any meaningful way that lightens the supervisor's core workload.
Augmentation potentialclaude-sonnet-53/5AI can assist supervisors with case documentation review, scheduling, tracking caseloads, or summarizing case notes, providing moderate productivity support while the supervisor retains full responsibility.
Task automatabilityclaude-haiku-4-5-202510011/5Supervising social workers requires real-time judgment about complex interpersonal dynamics, performance feedback, professional development, and case-specific guidance. No current AI system can end-to-end manage these supervisory relationships with the nuanced accountability required.
Task automatabilityclaude-sonnet-51/5Supervising staff involves relational trust-building, performance evaluation, mentorship, and judgment calls about clinical decisions that current AI cannot execute end-to-end.rated as no meaningful automation potential exists today.rated as low.
Adoption barriersclaude-haiku-4-5-202510015/5Supervision of licensed professionals is a core management and fiduciary responsibility. Employment law, professional licensing boards, and organizational liability structures explicitly require a licensed human supervisor to oversee other social workers' work.
Adoption barriersclaude-sonnet-55/5Supervision of licensed social workers typically requires a credentialed supervisor per state licensing boards and agency policy, creating a hard legal/professional barrier to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Implementing AI monitoring systems would still require a human supervisor to verify, respond to, and act on alerts—adding layers of cost rather than displacing the supervisor's loaded wage.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this task, so cost comparison favors the human entirely; any AI attempt would require full human oversight anyway.
Technical feasibility todayclaude-haiku-4-5-202510011/5While AI can generate reports or flag performance metrics, no deployed product reliably performs supervisory oversight of human social workers in production. Supervision inherently requires human authority, trust, and accountability.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs supervisory functions for social workers; this remains firmly outside current product capability.

Place children in foster or adoptive homes, institutions, or medical treatment centers.

0

CI 0–0 · exposure 0 · augmentation 38 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Child welfare agencies are traditionally slow-moving, publicly-funded, and regulatory-bound organizations. Adoption of AI for child placement decisions remains minimal; the sector prioritizes legal compliance and professional judgment over technological efficiency.
Sector adoption velocityclaude-sonnet-51/5Public social services and child welfare agencies are slow-adopting, resource-constrained, and heavily regulated, showing minimal AI deployment for core casework decisions.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist with organizing case files, matching criteria to available placements, or flagging risk factors, but the core task—making a placement decision—fundamentally requires human professional expertise and legal authority. Assistance is limited to preparation and information gathering.
Augmentation potentialclaude-sonnet-53/5AI can help match records, flag available placements, or summarize case histories, offering moderate assistance while the actual placement decision remains human-led.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires nuanced human judgment about child safety, family circumstances, legal implications, and welfare outcomes. No current AI system can autonomously determine placement decisions that have profound, lasting impacts on vulnerable children's lives.
Task automatabilityclaude-sonnet-51/5Placing children requires in-person assessment, judgment about safety and fit, legal proceedings, and relationship-building that AI cannot perform end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Child protective services and foster/adoptive placement are heavily regulated and typically require licensed social workers and court oversight. Legal liability for placement outcomes, mandatory human accountability, and child welfare regulations create near-absolute barriers to automation.
Adoption barriersclaude-sonnet-55/5Placement decisions involve licensed social workers, court oversight, child welfare law, and high liability for errors, making this a hard-barrier task requiring human authorization.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI cannot replace the human social worker's role here; the task fundamentally requires licensed professional judgment and legal accountability. Any AI system would function only as a support tool, not a cost substitute for the human decision-maker.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so cost comparison favors the human social worker who must be involved by law and practice.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs autonomous child placement decisions in production. While AI can assist with data aggregation and flagging cases, the legal and ethical responsibility for placement decisions remains exclusively with human social workers and courts.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product independently places children in foster/adoptive homes or institutions; this remains entirely a human, agency-driven process.

Recommend temporary foster care and advise foster or adoptive parents.

0

CI 0–0 · exposure 0 · augmentation 38 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Child welfare and social services are traditionally low-digitization sectors with strong human-contact requirements, regulatory oversight, and organizational conservatism. Adoption of autonomous AI in this domain remains minimal and faces cultural and legal resistance.
Sector adoption velocityclaude-sonnet-51/5Social services and child welfare agencies are a low-digitization, highly regulated sector with minimal AI adoption for case decision-making.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist social workers by organizing case documentation, flagging relevant information, or summarizing parent histories, but the core task—assessing family suitability and recommending placement—requires human judgment, legal authority, and accountability that limits meaningful augmentation.
Augmentation potentialclaude-sonnet-53/5AI can help with documentation, case note summarization, matching database searches, and drafting guidance materials, but the core judgment and advising remain human-led.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires deep judgment about child welfare, family dynamics, and legal/ethical considerations that current AI cannot perform autonomously. The recommendation of temporary foster care involves high-stakes decisions about child placement that demand human expertise, investigation, and accountability that AI systems cannot provide.
Task automatabilityclaude-sonnet-51/5This task requires nuanced case-specific judgment, home visits, interviews, and legal/ethical accountability for child welfare decisions that AI cannot perform end-to-end today.'
Adoption barriersclaude-haiku-4-5-202510015/5This task is heavily protected by licensing requirements (social work licenses), legal authority (child protective services mandates), liability asymmetry (wrong placements cause severe harm), and regulatory frameworks that require a licensed professional to legally make and sign off on foster care recommendations.
Adoption barriersclaude-sonnet-55/5Foster care placement decisions are legally regulated, require licensed social workers, court oversight, and carry high liability for child safety, creating hard barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of an AI system with sufficient integration, oversight, and liability coverage to handle foster care recommendations would far exceed the loaded cost of a human social worker, given the high-stakes nature and legal requirements of the task.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the human assessment and relationship-building required, so there is no meaningful cost comparison—human labor remains necessary.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs independent foster care placement recommendations or advisory services to adoptive/foster parents in production. These decisions remain the domain of licensed social workers and require investigation, interviews, and legal authority that AI systems do not possess.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product independently recommends foster placements or advises foster/adoptive parents; this remains squarely a human casework function.

Serve as liaisons between students, homes, schools, family services, child guidance clinics, courts, protective services, doctors, and other contacts to help children who face problems, such as disabilities, abuse, or poverty.

0

CI 0–0 · exposure 0 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Child welfare, family services, and school-based social work are highly regulated, human-centered sectors with slow digital transformation, minimal AI integration, and strong professional licensing requirements that limit automation.
Sector adoption velocityclaude-sonnet-51/5Social work in schools and child protective services is a low-digitization, high-touch, relationship-driven sector with minimal AI agent deployment in this specific liaison function.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could modestly assist with scheduling, coordinating contact information across agencies, or summarizing case notes, but the core liaison, advocacy, and judgment work cannot be meaningfully augmented by current AI systems without removing the human from the loop.
Augmentation potentialclaude-sonnet-53/5AI can help with case note summarization, scheduling, information retrieval across systems, and drafting communications, meaningfully aiding the administrative side of this liaison work.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires real-time relationship-building, complex judgment about child welfare and family dynamics, coordination across multiple stakeholders with conflicting interests, and navigation of highly sensitive situations where errors can cause serious harm. Current AI cannot perform the core liaison, advocacy, and trust-building functions that are essential to this work.
Task automatabilityclaude-sonnet-51/5This task requires in-person relationship building, trust, judgment about safety, and coordination across multiple human institutions—none of which can be automated end-to-end by current AI.
Adoption barriersclaude-haiku-4-5-202510015/5Hard legal and regulatory barriers protect this task: social workers must be licensed professionals, child welfare decisions require human accountability in court proceedings, mandatory reporter status carries legal liability, and protective services agencies require human-authorized decision-makers under state child welfare law.
Adoption barriersclaude-sonnet-55/5Child welfare cases often involve legal mandates, licensed social worker requirements, court testimony, and liability for abuse/neglect determinations, creating hard regulatory and professional barriers.
Cost vs. human wageclaude-haiku-4-5-202510011/5Fully deployed AI agents capable of coordinating across child welfare systems would require significant oversight, liability coverage, and integration costs that would exceed the cost of a social worker's time, especially given the consequence-sensitive nature of the task.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this liaison function, so cost comparison favors the human worker entirely; AI cannot deliver the output at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can reliably serve as a liaison between families, courts, schools, and protective services, as this requires legal authority, professional licensing, accountability for decisions affecting child safety, and human judgment about abuse, neglect, and family dynamics that courts and agencies must trust. This remains research-stage.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product acts as a liaison for at-risk children across schools, courts, and protective services; this remains firmly a human relational and advocacy role.

Counsel parents with child rearing problems, interviewing the child and family to determine whether further action is required.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Child welfare agencies remain among the least digitized sectors. Adoption of even basic case management tools is slow, and clinical judgment tasks are performed by licensed staff with no meaningful AI displacement in practice.
Sector adoption velocityclaude-sonnet-51/5Social services is a low-digitization, high-touch sector with minimal AI deployment in direct family counseling and risk assessment work.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with limited administrative tasks (documentation, scheduling) or as a preliminary screening tool, but meaningful augmentation is constrained by liability, the centrality of human relationship-building to counseling, and the need for professional judgment in safeguarding decisions.
Augmentation potentialclaude-sonnet-52/5AI can help with case note drafting, scheduling, or summarizing prior records, but offers little assistance during the actual sensitive interview and judgment process.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires nuanced human judgment, emotional intelligence, and clinical expertise to assess family dynamics and detect subtle signs of abuse or dysfunction. Current AI systems cannot reliably conduct therapeutic interviews, build trust, or make safeguarding decisions that meet the threshold for 50% time savings at equal quality.
Task automatabilityclaude-sonnet-51/5This requires live rapport-building, reading emotional cues, and clinical judgment about child welfare risk—AI cannot conduct these interviews or make safety determinations end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Licensing laws, mandatory reporter obligations, and liability requirements legally necessitate a qualified human social worker to conduct family interviews and make disposition decisions. Regulatory frameworks and child welfare law hard-bind this task to credentialed professionals.
Adoption barriersclaude-sonnet-55/5Licensed social workers are legally required to conduct these assessments, with mandatory reporting duties and liability for child safety determinations that cannot be delegated to software.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI inference cost is minimal, but the task requires professional oversight, liability coverage, and human clinician involvement for legal and ethical compliance, making total cost per intervention comparable to or exceeding direct human delivery.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so cost comparison favors the human by default; any AI attempt would carry unacceptable liability risk.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs clinical family counseling and abuse risk assessment. While chatbots exist, they cannot legally or ethically substitute for licensed social workers in this context, and no production system has demonstrated reliable capability for this safety-critical task.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs child welfare counseling interviews or risk determinations in production; this remains firmly a human clinical function.

Address legal issues, such as child abuse and discipline, assisting with hearings and providing testimony to inform custody arrangements.

0

CI 0–0 · exposure 0 · augmentation 38 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Child welfare and family law remain highly regulated, human-centered sectors with minimal AI adoption for core legal functions. These tasks sit at the intersection of law, child protection, and social work—domains resistant to automation.
Sector adoption velocityclaude-sonnet-51/5Social work and family court systems are highly regulated, low-digitization environments with strong human-centric requirements, showing minimal AI adoption for this specific legal/testimonial function.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might assist with case documentation, report writing, or evidence organization, but the core act of legal testimony and custody judgment requires deep human expertise and cannot be meaningfully augmented by current systems.
Augmentation potentialclaude-sonnet-53/5AI can assist with case documentation, evidence organization, drafting reports, and research to prepare for hearings, but cannot replace the professional's direct testimony or judgment during legal proceedings.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires legal judgment, courtroom presence, expert testimony, and human discretion in sensitive child welfare matters. Current AI cannot reliably perform legal representation, courtroom testimony, or make custody recommendations without a licensed human social worker.
Task automatabilityclaude-sonnet-51/5This requires in-person testimony, sworn statements, courtroom presence, and professional judgment tied to legal accountability; no current AI system can substitute for a social worker's testimony or legal representation in custody hearings.
Adoption barriersclaude-haiku-4-5-202510015/5Legal testimony in custody cases requires a licensed social worker; courts mandate sworn, accountable human experts. Liability, regulatory requirements, and the need for human judgment on child safety create hard legal barriers to automation.
Adoption barriersclaude-sonnet-55/5Legal testimony, custody determinations, and child abuse investigations require licensed professionals with legal standing, sworn oath obligations, and personal accountability that cannot be delegated to AI systems.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI would require extensive human oversight, legal review, and ultimately a licensed social worker to deliver the testimony anyway, making the all-in cost substantially higher than direct human performance of the task.
Cost vs. human wageclaude-sonnet-51/5AI cannot perform this task at all, so there is no viable cost comparison—human professional involvement is mandatory and non-substitutable.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs legal testimony or custody-arrangement consultation autonomously. While AI can assist with document drafting, the core task—appearing at hearings and providing expert judgment—remains exclusively human-performed in practice.
Technical feasibility todayclaude-sonnet-51/5No deployed product provides court testimony or represents social workers in legal proceedings; this remains entirely research-stage/non-existent for this specific function.

Lead group counseling sessions that provide support in such areas as grief, stress, or chemical dependency.

0

CI 0–0 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Group therapeutic practice is highly regulated, requires human licensure, and operates in sectors (mental health, schools, child welfare) with slow digital transformation and strong professional gatekeeping against algorithmic substitution.
Sector adoption velocityclaude-sonnet-51/5Social work and mental health services are a low-digitization, high-human-contact sector where AI adoption for direct clinical group facilitation remains essentially nonexistent in production.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could modestly assist by transcribing session notes or suggesting follow-up resources, but the core therapeutic work—facilitating dialogue, managing interpersonal dynamics, and making clinical judgments—remains irreducibly human, limiting meaningful augmentation.
Augmentation potentialclaude-sonnet-53/5AI can help social workers prepare session materials, summarize case notes, suggest discussion prompts, or draft psychoeducational content, but this is supportive rather than transformative to the live facilitation task.
Task automatabilityclaude-haiku-4-5-202510011/5Leading group counseling sessions requires sustained emotional intelligence, real-time responsiveness to individual psychological states, therapeutic alliance-building, and the ability to manage complex group dynamics—capabilities that current AI systems cannot reliably perform end-to-end. No current system can replace a licensed therapist facilitating therapeutic groups.
Task automatabilityclaude-sonnet-51/5Leading live group counseling requires real-time facilitation, reading emotional dynamics among multiple people, crisis intervention, and building trust—capabilities current AI cannot deliver end-to-end in a real group setting.
Adoption barriersclaude-haiku-4-5-202510015/5Social work is a licensed profession; only credentialed social workers, psychologists, or counselors can legally lead therapeutic group sessions in most jurisdictions. Liability, duty of care, and regulatory requirements create hard legal barriers to automation.
Adoption barriersclaude-sonnet-55/5Group counseling on sensitive topics like chemical dependency typically requires a licensed clinical social worker, mandated reporting duties, liability for participant safety, and legal/ethical accountability that cannot be delegated to AI.
Cost vs. human wageclaude-haiku-4-5-202510011/5A human social worker leading group sessions must be paid a professional wage ($45–70k annually loaded), and AI oversight or augmentation tools cost far less than replacing the core function, but full automation is not feasible, making direct cost comparison moot.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task, so cost comparison favors the human entirely; any attempted AI substitute would require extensive human oversight negating cost savings.
Technical feasibility todayclaude-haiku-4-5-202510011/5While chatbots can simulate supportive dialogue, no deployed product reliably conducts actual group counseling sessions with the clinical judgment, ethical responsibility, and therapeutic presence required. Production systems do not exist for this task.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently leads group therapy or counseling sessions for grief, stress, or chemical dependency; existing AI chatbots offer only individual, narrow-scope supplementary support, not group facilitation.

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.