Probation Officers and Correctional Treatment Specialists

21-1092.00

Median wage $66,270/yr89,390 employed (US)Rank #770 of 923 scored · top 83% by substitution

Provide social services to assist in rehabilitation of law offenders in custody or on probation or parole. Make recommendations for actions involving formulation of rehabilitation plan and treatment of offender, including conditional release and education and employment stipulations.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution16

Exposure16

Augmentation47

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

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

Why this score

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

Task automatabilityw 35%18

panel mean rating 1.7/5 → substitution pressure 18/100

Technical feasibility todayw 20%14

panel mean rating 1.6/5 → substitution pressure 14/100

Cost vs. human wagew 15%21

panel mean rating 1.8/5 → substitution pressure 21/100

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%10

panel mean rating 1.4/5 → substitution pressure 10/100

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.

Develop and prepare packets containing information about social service agencies, assistance organizations, and programs that might be useful for inmates or offenders.

CI 34–65 · exposure 50 · augmentation 75 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Correctional systems are slow digital adopters with limited IT infrastructure and risk-averse organizational cultures. While some jurisdictions pilot case management tools, production deployment of automated resource packet generation remains rare in the corrections sector.
Sector adoption velocity	claude-sonnet-5	2/5	Corrections and probation agencies are typically slow adopters of AI tools due to limited budgets, legacy systems, and cautious public-sector IT policies.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can substantially assist probation officers by rapidly surfacing relevant social service agencies, filtering programs by eligibility criteria, and drafting initial packets, allowing officers to focus on personalizing recommendations and case-specific judgment rather than spending time on research and data compilation.
Augmentation potential	claude-sonnet-5	4/5	AI can significantly speed up drafting, updating, and formatting these resource packets, letting officers focus on verifying and tailoring content rather than manual compilation.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can gather and compile information about social service agencies and programs, assembling contextualized packets tailored to individual inmate/offender needs requires human judgment about relevance, appropriateness, and local resources. Current systems can accelerate research and data collection but cannot reliably determine which resources match specific client circumstances without substantial human review.
Task automatability	claude-sonnet-5	4/5	Compiling and organizing information about social service agencies and programs into reference packets is largely a research-and-formatting task that current AI can do well, given directories or web access, with human review for accuracy and local relevance.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Correctional and probation systems are highly regulated, risk-sensitive environments where inaccurate or inappropriate resource guidance carries liability and public safety implications. Organizational culture typically requires human professional judgment and accountability in decisions affecting offender reintegration, creating strong adoption friction.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human create these packets, but agencies may want oversight to ensure accuracy and appropriateness of referrals for offenders, creating mild organizational friction.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	The cost of AI-assisted information gathering and document assembly is roughly comparable to probation officer labor when accounting for integration, verification overhead, and the need for human review to ensure appropriateness and accuracy of resource recommendations.
Cost vs. human wage	claude-sonnet-5	4/5	Generating and updating informational packets via AI is far cheaper than staff hours spent researching and compiling resource directories, though verification of local program details still requires some human time.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Information retrieval and document compilation tools exist and are increasingly used in administrative contexts, but production deployments for correctional settings remain limited. Existing systems can draft packets with material accuracy issues and struggle with currency of resource listings, requiring significant human oversight and verification.
Technical feasibility today	claude-sonnet-5	3/5	AI writing/research tools can draft such compilations today, but no widely deployed correctional-sector product specifically automates this task in production; most agencies still rely on manually maintained resource lists.

Prepare and maintain case folder for each assigned inmate or offender.

CI 25–60 · exposure 45 · augmentation 63 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Correctional systems are typically low-digitization, slower-moving organizations with legacy infrastructure and conservative practices. AI adoption in case management remains at the pilot stage; production deployment of autonomous case folder management is rare in the sector.
Sector adoption velocity	claude-sonnet-5	2/5	Public sector corrections and probation agencies typically lag in technology adoption due to budget constraints, legacy systems, and cautious approaches to sensitive offender data.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist by drafting narrative summaries, flagging inconsistencies in submitted documents, and organizing information—raising efficiency on routine documentation tasks. However, the human officer must retain full judgment authority over content, redaction, and legal sufficiency, so augmentation is partial rather than transformative.
Augmentation potential	claude-sonnet-5	4/5	AI can meaningfully assist officers by auto-populating forms, summarizing case histories, and flagging missing documentation, significantly speeding up file preparation and maintenance while the officer retains oversight.
Task automatability	claude-haiku-4-5-20251001	2/5	While document generation and basic record compilation can be partially automated, case folders require ongoing judgment about relevance, sensitive redaction decisions, and integration of complex behavioral/legal updates. Current AI cannot reliably manage the full lifecycle without substantial human oversight, falling short of 50% time savings at equal quality.
Task automatability	claude-sonnet-5	4/5	Preparing and maintaining case folders is largely structured document assembly, data entry, and record organization, which AI/document-management systems can substantially automate given standardized templates and inputs from case management systems.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: case folders are legal documents often used in court proceedings, requiring certified personnel sign-off; liability is asymmetric (errors harm offenders and public safety); and regulations typically mandate that a licensed probation officer maintain official accountability for record accuracy and completeness.
Adoption barriers	claude-sonnet-5	3/5	Case files often contain legally sensitive and confidential information requiring accountability of a designated officer, and agencies may have policies mandating human recordkeeping responsibility, creating moderate procedural and privacy-related barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration and oversight costs for AI-assisted case folder management remain high relative to a probation officer's loaded wage, especially given liability and accuracy requirements. The cost advantage, if any, is modest and offset by error-correction overhead.
Cost vs. human wage	claude-sonnet-5	4/5	Automated document generation and record-keeping tools are inexpensive per case relative to officer time spent on clerical file maintenance, though oversight and correction of errors add some cost.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably handles correctional case folder management end-to-end in production. Some case management systems exist with document automation features, but these are narrow tools requiring significant human oversight and are not mature solutions in actual correctional facilities.
Technical feasibility today	claude-sonnet-5	3/5	Case management software with automated form population and record-keeping exists and is used in corrections/probation agencies, but full end-to-end folder maintenance still requires human review, verification of legal accuracy, and integration with government-specific systems, limiting fully autonomous deployment.

Recommend appropriate penitentiary for initial placement of an offender.

CI 25–55 · exposure 45 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Corrections agencies and probation departments are typically risk-averse, underfunded, and slower to adopt AI than commercial sectors. Pilot programs exist, but production deployment of autonomous placement systems remains limited due to legal and reputational concerns.
Sector adoption velocity	claude-sonnet-5	2/5	Corrections is a historically slow-adopting, highly regulated public-sector field with limited digitization and cautious integration of automated decision tools.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI-powered placement tools can significantly assist officers by surfacing relevant offender data, flagging institutional constraints, and suggesting matches, allowing the human to focus on nuanced judgment and accountability rather than data gathering and routine matching.
Augmentation potential	claude-sonnet-5	3/5	AI-driven risk assessment and classification tools can meaningfully inform and speed up an officer's evaluation process, even though the final recommendation remains human-driven.
Task automatability	claude-haiku-4-5-20251001	4/5	AI systems can analyze offender characteristics (criminal history, offense type, security risk scores, behavioral assessments) against institutional criteria to recommend placement with high consistency. While some discretionary judgment remains, the core matching task is largely rule-based and data-driven, achieving significant time savings over manual review.
Task automatability	claude-sonnet-5	2/5	AI can assist in aggregating risk factors and classification data, but final placement recommendations require judgment integrating legal, safety, and case-specific nuances that current systems cannot fully replicate end-to-end at equal quality.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Legal and liability barriers are substantial: placement decisions directly affect an offender's liberty and safety, and an incorrect recommendation could expose the agency to civil liability. Professional judgment and human accountability are legally and organizationally expected, creating resistance to full automation.
Adoption barriers	claude-sonnet-5	4/5	Placement decisions affect liberty, safety, and legal rights, typically requiring sign-off by a qualified corrections official; institutional and legal frameworks mandate human authorization for such consequential decisions.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Implementing and running an AI-based placement system costs far less than the time a probation officer spends manually researching offender profiles and institutional capacities. Inference and integration costs are minimal relative to the loaded wage of a specialized corrections professional.
Cost vs. human wage	claude-sonnet-5	2/5	Risk assessment software has modest licensing costs but requires human oversight, data entry, and review, so total cost is not dramatically cheaper than officer time, though some efficiency gains exist.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Decision-support systems and risk-assessment algorithms exist in corrections (e.g., COMPAS, facility-matching tools), but fully autonomous placement recommendation without human review is rare in production. Most deployed systems serve as decision aids requiring human sign-off rather than autonomous end-to-end systems.
Technical feasibility today	claude-sonnet-5	2/5	Some correctional systems use risk-assessment software (e.g., classification instruments) to inform placement, but these are decision-support tools, not autonomous recommenders; humans finalize the decision in production settings.

Write reports describing offenders' progress.

CI 25–37 · exposure 33 · augmentation 63 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Criminal justice agencies are characteristically slow to digitize and adopt AI, with high organizational and political friction around automation of sensitive population management. Few jurisdictions have moved beyond pilots; laggard-sector patterns dominate (small budgets, legacy systems, public sector inertia).
Sector adoption velocity	claude-sonnet-5	2/5	Government/criminal justice sectors are slow adopters of AI due to public accountability, data sensitivity, and procurement constraints, with pilots rare and production use rarer still.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist by auto-generating chronological summaries, extracting key events from intake files, and suggesting standard language for routine updates, which would accelerate the drafting phase. However, the human officer must still synthesize, verify, interpret, and take legal responsibility for the final product.
Augmentation potential	claude-sonnet-5	4/5	AI can meaningfully speed up drafting of routine narrative sections, summarizing case history and progress notes, while the officer retains responsibility for accuracy and final content.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can draft factual summaries of recorded events and generate template-based progress notes, the task requires substantive judgment about behavioral change, risk assessment, and treatment efficacy—elements that demand human observation and professional interpretation. Current systems cannot reliably synthesize complex case histories into legally defensible progress documentation without extensive manual review.
Task automatability	claude-sonnet-5	3/5	AI can draft progress report narratives from case notes and structured data, but requires human input, verification, and judgment calls, limiting full end-to-end automation.5
Adoption barriers	claude-haiku-4-5-20251001	4/5	Reports must be signed by the probation officer and are discoverable legal documents; errors can affect parole, sentencing, and litigation. Regulatory frameworks and professional accountability standards effectively require a licensed human to author and attest to the content, creating a strong legal and liability barrier to full automation.
Adoption barriers	claude-sonnet-5	4/5	These reports carry legal weight (parole decisions, court records) and typically require a certified officer's sign-off, involve confidential criminal justice data, and are subject to agency liability and record-integrity rules.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The cost of AI-assisted report writing (inference plus required human review and correction) approaches or exceeds the marginal time saved, especially when error costs and liability exposure are factored in. A probation officer's hourly wage is competitive with the all-in cost of a system requiring significant oversight.
Cost vs. human wage	claude-sonnet-5	3/5	AI drafting could cut time substantially, but oversight, data entry into secure systems, and verification against confidential records keep overall cost savings moderate rather than order-of-magnitude.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No production system reliably writes criminal justice progress reports autonomously; existing document generation tools are narrow in scope and require heavy human oversight. Jurisdictions have piloted AI for risk assessment and summary drafting, but deployed products consistently require probation officers to substantially rewrite or validate AI-generated content.
Technical feasibility today	claude-sonnet-5	2/5	General-purpose LLMs can draft such text, but no widely deployed, validated product specifically integrated into probation case management systems performs this reliably at scale today.

Arrange for postrelease services, such as employment, housing, counseling, education, and social activities.

CI 25–25 · exposure 25 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Correctional and probation services are traditionally low-tech, fragmented across public agencies with limited digitization. While some jurisdictions are piloting case management tools, broad AI adoption for arranging postrelease services remains nascent; most systems still rely on manual coordination.
Sector adoption velocity	claude-sonnet-5	2/5	Corrections and social services is a historically slow-adopting, resource-constrained public sector with limited AI deployment for case coordination tasks, resulting in low actual adoption despite some pilot programs.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist by surfacing available housing, employment, and counseling resources; flagging client eligibility for programs; and drafting communication templates. However, the human probation officer must still navigate provider availability, negotiate slots, and adapt plans based on real-world friction.
Augmentation potential	claude-sonnet-5	3/5	AI can meaningfully assist with tasks like searching for available resources, drafting referral letters, and organizing information, improving efficiency while the officer still manages relationships and decisions.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can help identify available resources and draft referral letters, arranging services requires navigating complex eligibility criteria, building relationships with service providers, and making individualized judgment calls about client needs. Current systems cannot reliably complete this end-to-end task at 50% time savings; much requires human negotiation and follow-up.
Task automatability	claude-sonnet-5	2/5	Coordinating and arranging services requires relationship-building, negotiation with local providers, and judgment about individual client needs that current AI cannot fully replicate end-to-end., though it can assist with parts like drafting referrals or searching resource databases.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Probation officers and treatment specialists are often licensed professionals whose authority and discretion in case management are legally and ethically bound. Clients often require individualized assessment and trust-building; many jurisdictions have licensing/oversight requirements for who can authorize referrals and manage client welfare.
Adoption barriers	claude-sonnet-5	4/5	This task often involves statutory obligations, case management accountability, and requires a licensed/certified officer to make judgment calls and maintain liability for client outcomes, creating substantial regulatory and organizational barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for resource lookup and administrative tasks are inexpensive, but the bulk of the loaded cost remains human caseworker time spent on relationship-building, negotiation, and problem-solving when placements fall through. AI cost per successful arrangement is not yet an order of magnitude lower.
Cost vs. human wage	claude-sonnet-5	2/5	While AI could cheaply assist with research and drafting communications, the human labor of building relationships with employers, landlords, and service providers plus follow-up cannot be replaced cheaply, keeping overall costs comparable to human labor.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably arranges multi-service postrelease coordination autonomously. Some CMS tools assist with resource databases or documentation, but actual arrangement—contacting providers, confirming slots, handling rejections—remains manual and human-driven in correctional practice.
Technical feasibility today	claude-sonnet-5	2/5	There are no deployed products specifically automating the arrangement of postrelease services; general-purpose AI tools could help with research and communication but no mature production system performs this task reliably in corrections agencies.

Provide offenders or inmates with assistance in matters concerning detainers, sentences in other jurisdictions, writs, and applications for social assistance.

CI 18–25 · exposure 20 · augmentation 50 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Correctional systems are traditionally low-digitization sectors with slow AI adoption, limited funding for new tools, and heavy reliance on established personnel practices. While some jurisdictions pilot case management systems, deep adoption of AI-driven legal assistance in corrections remains nascent.
Sector adoption velocity	claude-sonnet-5	1/5	Corrections and probation is a slow-adopting, highly regulated, public-sector field with limited digitization and cautious AI uptake compared to fast-moving sectors like finance or tech.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist probation officers by automating literature searches across jurisdictions, drafting initial applications, and flagging relevant statutes or program eligibility criteria, thereby reducing research time and ensuring fewer opportunities are missed. However, the human must verify all legal claims and tailor advice to individual circumstances.
Augmentation potential	claude-sonnet-5	3/5	AI tools can help officers research applicable laws across jurisdictions, draft forms, and summarize case files, providing meaningful assistance while the officer retains responsibility for judgment and communication.
Task automatability	claude-haiku-4-5-20251001	2/5	This task requires navigating complex legal frameworks, understanding jurisdiction-specific sentencing rules, and providing personalized guidance based on individual offender circumstances. While AI could assist with information retrieval and document drafting, the need for contextual judgment, legal accuracy, and individualized counsel prevents full automation without substantial human oversight.
Task automatability	claude-sonnet-5	2/5	This requires navigating multiple legal jurisdictions, interpreting case-specific documents, and providing personalized guidance that combines legal knowledge with case management judgment, limiting full automation.but AI could assist with research and drafting.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Legal liability, correctional system oversight requirements, and the need for human judgment in matters affecting incarceration and release create significant barriers. Many jurisdictions require licensed or trained personnel to provide legal guidance, and errors carry high consequences for both offenders and public safety.
Adoption barriers	claude-sonnet-5	4/5	This involves legal/administrative matters with liability implications, often requiring a credentialed probation officer to interpret and act on legal documents and coordinate with courts and agencies, creating substantial institutional and regulatory barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The cost of AI legal research and document generation tools, combined with mandatory human review of output for accuracy and liability reasons, approaches or exceeds the cost of a probation officer providing the service directly, especially given error-correction overhead.
Cost vs. human wage	claude-sonnet-5	2/5	AI could reduce time on document lookup and drafting, but the human oversight, legal liability, and case-specific judgment required keep costs comparable to human labor rather than dramatically cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably handles the full scope of multi-jurisdictional legal research, sentence interpretation, and social assistance application guidance at production scale. Tools exist for legal research and document generation, but they lack the integrated capability to synthesize jurisdiction-specific detainer rules, sentencing variations, and social program eligibility criteria with reliable accuracy.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs end-to-end coordination of detainers, cross-jurisdictional sentencing issues, and social assistance applications for offenders in a production correctional setting today.

Inform offenders or inmates of requirements of conditional release, such as office visits, restitution payments, or educational and employment stipulations.

CI 0–37 · exposure 20 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Correctional and probation systems are slow-moving, highly regulated, and risk-averse. Adoption of AI for core offender-facing communication remains minimal, with agencies prioritizing human judgment and legal compliance over automation.
Sector adoption velocity	claude-sonnet-5	2/5	Corrections agencies are historically slow adopters of AI due to public-sector budget constraints, unionized labor, and cautious risk posture around offender management.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with drafting standard documentation or preparing information summaries before an officer meets with an offender, but the direct communication task itself leaves limited room for augmentation, as the officer must personally deliver and validate understanding.
Augmentation potential	claude-sonnet-5	4/5	AI can help officers draft consistent, clear explanations of conditions, generate personalized checklists, and produce multilingual materials, meaningfully aiding the human-delivered task.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires real-time, personalized communication with individuals in the criminal justice system, with legal and binding consequences. While AI could draft standardized information, the actual delivery and explanation of conditional release terms demands human judgment, legal authority, and the ability to address individual questions or concerns—capabilities current AI systems cannot reliably perform end-to-end.
Task automatability	claude-sonnet-5	3/5	Generating and communicating standardized release conditions is a templated information-delivery task that AI could draft or explain, but confirming understanding, answering individualized questions, and legal accountability limit full automation.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Hard legal and regulatory barriers exist: a licensed probation officer must inform offenders of their release conditions and explain their obligations, as this communication has legal standing and enforcement consequences. Liability for miscommunication or failure to properly inform is borne by the officer and agency.
Adoption barriers	claude-sonnet-5	4/5	This is a legally significant notification tied to due process and supervision compliance, typically requiring a certified officer to deliver and document it, creating substantial regulatory and liability barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of deploying AI systems with adequate oversight and integration into correctional workflows would exceed the loaded wage of a probation officer for this task, especially when factoring in liability management and human review.
Cost vs. human wage	claude-sonnet-5	3/5	AI-generated notices or automated reminder systems could be cheap to produce, but human officer time is still required for verification, questions, and legal sign-off, keeping costs comparable rather than dramatically lower.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably performs the actual communication of legal release conditions to offenders or inmates. This requires authorized human interaction with individuals subject to legal supervision, and no current system operates autonomously in this capacity in production.
Technical feasibility today	claude-sonnet-5	2/5	Chatbots and document-generation tools can produce release condition summaries, but no widely deployed product in corrections systems handles formal offender notification end-to-end today.

Identify and approve work placements for offenders with community service sentences.

CI 13–25 · exposure 20 · augmentation 38 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Criminal justice and corrections systems are among the slowest adopters of automation; legacy systems, risk-averse organizational culture, and union agreements limit deployment of autonomous decision-making, even when technical capability exists.
Sector adoption velocity	claude-sonnet-5	1/5	Corrections and probation agencies are a low-digitization, government sector with minimal AI adoption for case-specific discretionary decisions.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist by identifying and pre-screening suitable placements, flagging eligibility issues, and organizing job-matching data, allowing the officer to focus review on suitability and approval. However, the core judgment and authority remain human-centered.
Augmentation potential	claude-sonnet-5	2/5	AI could help search databases of approved worksites or flag scheduling conflicts, offering modest assistance, but does not transform the core judgment-based placement task.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist in matching offenders to available placements based on skills and sentence requirements, the task requires nuanced judgment about individual suitability, risk assessment, and community integration that current systems cannot reliably perform end-to-end. Final approval decisions involve legal and safety considerations that remain heavily human-dependent.
Task automatability	claude-sonnet-5	2/5	Matching offenders to appropriate community service placements requires judgment about risk, safety, offender skills, and community relationships that current AI cannot reliably execute end-to-end.dana ArationaleAI could assist with matching/logistics but not fully replace the decision.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Approval of work placements for offenders carries significant regulatory and legal weight; supervisory authority and sign-off by a licensed probation officer or specialist are typically required by statute or agency policy, creating a hard barrier to full automation.
Adoption barriers	claude-sonnet-5	5/5	Approval of community service placements is a legally mandated function tied to court orders and supervision authority, requiring a licensed/authorized officer to sign off, creating strong regulatory and liability barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI screening and matching tools have modest deployment costs, but the oversight, validation, and human review required to ensure appropriate placement approval—including liability exposure—keep total cost near or above that of human specialists performing the task.
Cost vs. human wage	claude-sonnet-5	2/5	Even if AI assisted with matching, human review, site vetting, and legal approval remain necessary, keeping overall cost closer to human-driven processes with only marginal savings.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed system reliably performs approval of work placements autonomously; existing tools may support matching or scheduling but lack the contextual judgment and legal authority to approve placements. Products exist for eligibility screening or job matching, but they require substantial human review and decision-making.
Technical feasibility today	claude-sonnet-5	1/5	No deployed products perform this specific placement-approval task in corrections systems; case management software may track data but does not autonomously identify/approve placements.

Develop rehabilitation programs for assigned offenders or inmates, establishing rules of conduct, goals, and objectives.

CI 9–25 · exposure 17 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Corrections and probation agencies are typically low-digitization, risk-averse government organizations with limited AI adoption; cultural and regulatory resistance to algorithmic decision-making in criminal justice remains strong.
Sector adoption velocity	claude-sonnet-5	2/5	Corrections and probation agencies are historically slow adopters of new technology, constrained by budgets, unions, and legal/regulatory caution, so AI tools remain in pilot stages.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist probation officers by organizing offender data, flagging risk factors, suggesting evidence-based program components, and drafting program templates, thereby supporting faster and more consistent program development while the officer retains decision authority.
Augmentation potential	claude-sonnet-5	3/5	AI can help officers draft goal language, summarize case histories, or suggest evidence-based intervention options, meaningfully speeding up parts of plan preparation while the officer retains final judgment.
Task automatability	claude-haiku-4-5-20251001	1/5	Developing individualized rehabilitation programs requires nuanced judgment about offender circumstances, criminogenic needs, and tailored interventions that depend on human expertise and ethical accountability. Current AI cannot reliably perform this end-to-end with equivalent quality.
Task automatability	claude-sonnet-5	2/5	AI can draft template rehabilitation plans or suggest goals based on case data, but individualized program design requires clinical judgment, risk assessment nuance, and accountability that current systems cannot fully replicate end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Correctional and probation work is heavily regulated; rehabilitation program development must be performed or signed off by licensed professionals, and liability for inadequate programs rests with human specialists, creating hard legal and organizational barriers to automation.
Adoption barriers	claude-sonnet-5	4/5	This is a statutorily assigned duty tied to public safety and legal accountability; supervision plans typically require a certified probation officer's judgment and sign-off, creating strong professional and liability barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for document drafting and data analysis are relatively inexpensive, but integration into correctional workflows and required human oversight still necessitate professional labor, making the all-in cost comparable to or higher than human-only approaches.
Cost vs. human wage	claude-sonnet-5	2/5	AI drafting tools are cheap per use, but the human officer must still review, personalize, verify legal compliance, and take responsibility, keeping overall cost savings modest.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While AI can assist with data synthesis and template generation, no deployed system reliably develops complete rehabilitation programs independently; human specialists must conduct assessments and make final determinations given liability and legal requirements.
Technical feasibility today	claude-sonnet-5	2/5	Some case management software includes risk-assessment tools and templated plan generation, but no deployed product autonomously develops full rehabilitation programs used reliably at scale without officer authorship.

Gather information about offenders' backgrounds by talking to offenders, their families and friends, and other people who have relevant information.

CI 4–23 · exposure 13 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Corrections and probation are heavily regulated, human-centric sectors with slow digital transformation. There is minimal evidence of AI adoption for core investigative or interview functions in probation departments.
Sector adoption velocity	claude-sonnet-5	1/5	Corrections and probation is a public-sector, low-digitization field with minimal AI agent deployment for casework interviews, lagging far behind sectors like finance or tech.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI might assist by transcribing interviews, organizing notes, or flagging patterns in existing data, but offers limited augmentation for the core task of conducting interviews and gathering sensitive background information through conversation.
Augmentation potential	claude-sonnet-5	3/5	AI can help by transcribing interviews, summarizing background records, and drafting reports, giving moderate assistance while the human conducts the actual information-gathering conversations.
Task automatability	claude-haiku-4-5-20251001	1/5	Gathering sensitive personal information from multiple sources requires real-time interpersonal judgment, rapport-building, and trust that current AI systems cannot establish. The task fundamentally depends on human-to-human interviews where subjects may only open up to actual people, making end-to-end automation infeasible.
Task automatability	claude-sonnet-5	2/5	The core activity is conducting sensitive interpersonal interviews requiring trust-building, judgment about credibility, and reading emotional cues, which current AI cannot substitute for in the interview itself, though transcription/summarization can save some time.rea
Adoption barriers	claude-haiku-4-5-20251001	5/5	This task is legally and ethically constrained: probation officers are mandated professionals required by law to conduct background investigations and have legal authority to compel information. Only licensed officers can legally conduct certain interviews and file official reports.
Adoption barriers	claude-sonnet-5	4/5	This is a sensitive, legally significant task tied to case investigations and court decisions, requiring trained, often government-employed officers with discretion and accountability, creating strong procedural and trust barriers to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI might assist with data organization and note-taking, but the core interviewing and relationship work requires a human probation officer. The full task cost remains dominated by human labor, with minimal cost displacement possible.
Cost vs. human wage	claude-sonnet-5	2/5	Since AI cannot substitute for the interviewing itself, the human labor cost remains largely unchanged; only marginal savings come from AI-assisted documentation, so overall cost ratio is not favorable to AI replacement.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product conducts multi-party interviews or background investigations autonomously. While AI can help organize information, it cannot reliably conduct interviews with offenders, families, or collateral contacts that yield accurate, actionable data.
Technical feasibility today	claude-sonnet-5	2/5	No deployed product conducts these sensitive interviews with offenders and family members; existing tools only assist with note-taking or transcription after a human has done the interpersonal work.

Arrange for medical, mental health, or substance abuse treatment services according to individual needs or court orders.

CI 4–23 · exposure 13 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Correctional and probation systems are traditionally low-digitization, bureaucratic sectors with high regulatory scrutiny and minimal automation; adoption of AI for placement decisions remains negligible in production.
Sector adoption velocity	claude-sonnet-5	1/5	Criminal justice and corrections agencies are notoriously slow adopters of AI due to funding constraints, legal liability concerns, and lack of digitization in case management systems.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI could assist by suggesting available treatment facilities, parsing court documents, or organizing client records, but the core task—matching individual needs to services and ensuring compliance—remains the officer's responsibility and judgment call.
Augmentation potential	claude-sonnet-5	3/5	AI tools can help officers search for available treatment programs, generate referral letters, and track compliance deadlines, offering meaningful assistance even though the human retains full responsibility for arranging services.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires navigating complex legal frameworks, assessing individual circumstances, coordinating with multiple service providers, and ensuring compliance with court orders—all judgment-intensive activities that current AI cannot perform end-to-end at quality parity with human specialists.
Task automatability	claude-sonnet-5	2/5	AI can help identify providers and draft referral paperwork, but arranging services requires judgment about individual circumstances, coordination with courts, and relationship-building with providers that current systems cannot fully execute end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong legal barriers exist: probation and correctional decisions are typically made by licensed professionals bound by statute; court orders must be documented and verified; liability for incorrect treatment placement falls on the officer, creating a hard requirement for human sign-off and discretion.
Adoption barriers	claude-sonnet-5	4/5	This task is tied to court orders and involves legal/statutory responsibility, confidentiality (health/substance abuse records), and accountability that typically requires a licensed officer to authorize and coordinate.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The cost of AI oversight, verification, and human review would likely approach or exceed the cost of a specialist directly arranging services, given the legal liability and accuracy requirements inherent in correctional contexts.
Cost vs. human wage	claude-sonnet-5	2/5	AI could cheaply assist with paperwork and provider lookup, but the human officer must still make judgment calls, communicate with courts and agencies, and take legal responsibility, so overall cost savings are modest.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably handles the full task of arranging court-ordered treatment services; AI can assist with database searches or scheduling, but the legal accountability, provider coordination, and individualized assessment remain human-dependent in practice.
Technical feasibility today	claude-sonnet-5	1/5	There are no deployed products that autonomously arrange court-mandated treatment services; this remains a human case-management function performed by officers.

Administer drug and alcohol tests, including random drug screens of offenders, to verify compliance with substance abuse treatment programs.

CI 0–25 · exposure 13 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Criminal justice and corrections remain relatively low-tech sectors with strong institutional conservatism, union protections, and regulatory compliance requirements; digital tools are piloted but deep automation is rare and faces resistance.
Sector adoption velocity	claude-sonnet-5	1/5	Corrections and probation work is a low-digitization, physically-grounded sector with minimal AI agent deployment for hands-on compliance tasks.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI could usefully assist by analyzing test results, flagging patterns of non-compliance, recommending next steps, and managing scheduling, raising officer productivity in record-keeping and analysis while the officer retains decision authority.
Augmentation potential	claude-sonnet-5	2/5	AI can help schedule random tests, track results, and flag anomalies in testing data, but offers no assistance with the physical administration itself.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI could theoretically assist with result interpretation and pattern analysis, the physical administration of tests (collection, chain-of-custody procedures) and real-time judgments about testing protocols require human presence and authority; no current AI system achieves 50% time savings end-to-end.
Task automatability	claude-sonnet-5	1/5	This requires physical specimen collection (urine, saliva, blood) and chain-of-custody handling, which AI cannot perform; it is an inherently physical, hands-on task.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Legal authority and liability are significant barriers: probation officers are court-appointed officials; a human officer must legally administer and document tests, verify identity, maintain chain-of-custody, and make compliance determinations that could affect incarceration decisions.
Adoption barriers	claude-sonnet-5	5/5	Legal chain-of-custody, evidentiary standards for court/parole compliance, and authorization requirements mean a qualified officer must personally administer and certify these tests.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI systems for result analysis and scheduling are relatively cheap, but they cannot replace the probation officer who must be physically present to administer tests and make judgment calls, so total cost savings are minimal relative to loaded human wages.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute for physical specimen collection, so cost comparison favors the human doing the physical task entirely.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some digital workflow tools exist to record test results and flag anomalies, but the core task of administering physical tests and making compliance decisions remains dependent on human staff; no deployed product reliably performs the full task autonomously.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product administers physical drug/alcohol tests; testing devices exist but require human administration and oversight.

Assess the suitability of penitentiary inmates for release under parole and statutory release programs and submit recommendations to parole boards.

CI 4–20 · exposure 13 · augmentation 50 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Correctional and parole systems are conservative, late-adopter sectors with deep regulatory and union constraints. Adoption of AI tools is slow and primarily in supporting data analysis rather than autonomous decision-making; no significant displacement of probation officers is evident in practice.
Sector adoption velocity	claude-sonnet-5	1/5	Corrections agencies are notoriously slow adopters of AI due to legal risk, public accountability concerns, and limited digitization of case management systems.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist by organizing case files, summarizing histories, and flagging statistical risk factors, improving efficiency and consistency in the preparation of assessments. However, the human probation officer remains the primary evaluator; augmentation is valuable but incremental.
Augmentation potential	claude-sonnet-5	3/5	AI-based risk assessment instruments and document summarization can meaningfully speed up file review and highlight relevant history, aiding but not replacing officer judgment.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires deep human judgment about rehabilitation, risk assessment, and individual circumstances; it involves synthesizing complex behavioral, psychological, and social factors that current AI cannot reliably evaluate end-to-end. No current system can autonomously conduct the multidimensional assessment and produce credible parole recommendations without substantial human oversight.
Task automatability	claude-sonnet-5	2/5	AI can synthesize case files and flag risk factors, but the holistic judgment integrating institutional behavior, victim input, and rehabilitation credibility resists full automation to a 50% time-saving-at-equal-quality bar today.a
Adoption barriers	claude-haiku-4-5-20251001	5/5	Parole and release decisions are legally mandated to involve human expert judgment and parole board review; jurisdictions have explicit regulatory and statutory requirements that a qualified human (probation officer, board member) must assess and recommend. Liability and the high cost of erroneous release create strong organizational and legal barriers to full automation.
Adoption barriers	claude-sonnet-5	5/5	Statutory and correctional regulations require a qualified officer to assess and formally recommend release, with legal accountability and due-process implications that preclude automated sign-off.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI-assisted tools may reduce administrative time, but the core task—expert human assessment—cannot be fully replaced. Integration costs, validation, and mandatory human review mean total cost savings are modest, and the loaded wage of a probation officer remains a significant baseline.
Cost vs. human wage	claude-sonnet-5	2/5	AI could cheaply draft summaries or flag risk indicators, but the human officer must still conduct interviews, verify records, and exercise judgment, keeping overall cost comparable to current staffing.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While AI tools can assist with data aggregation and flagging risk factors, no deployed product reliably performs the full assessment and recommendation task. Parole decisions involve sensitive legal and ethical judgments; existing risk-assessment tools have documented bias issues and are explicitly designed as decision-support aids, not autonomous decision-makers.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously produces parole suitability recommendations; risk-assessment tools exist but are decision-support inputs reviewed entirely by human officers.

Conduct prehearing and presentencing investigations and testify in court regarding offenders' backgrounds and recommended sentences and sentencing conditions.

CI 0–15 · exposure 8 · augmentation 38 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Correctional agencies operate under strict regulatory frameworks with slow digital adoption; human discretion and accountability are core requirements. No measurable displacement of this task by AI in production.
Sector adoption velocity	claude-sonnet-5	2/5	Criminal justice and corrections agencies are typically slow, under-resourced, and cautious adopters of AI, especially for functions carrying due-process and liability implications.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI might assist with document retrieval, case history summarization, or drafting report sections, but the core functions—investigation judgment, sentencing recommendations, and courtroom testimony—remain firmly human responsibilities. Assistance is marginal.
Augmentation potential	claude-sonnet-5	3/5	AI tools can help organize background records, draft report sections, and summarize risk-assessment data, meaningfully aiding preparation even though the investigation and testimony remain human-led.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires human judgment, legal expertise, and courtroom testimony—functions that are fundamentally tied to human authority and accountability. AI cannot independently conduct legally binding investigations or testify under oath in court.
Task automatability	claude-sonnet-5	2/5	AI can help draft parts of reports and summarize case files, but the core investigative work (interviews, home visits, risk judgment) and courtroom testimony require human presence and accountability that cannot be automated end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong legal barriers exist: only licensed probation officers are authorized to conduct official presentencing investigations, and testimony in court requires a human who can be sworn, cross-examined, and held accountable. Liability and evidentiary rules create hard barriers.
Adoption barriers	claude-sonnet-5	5/5	Courtroom testimony and official sentencing recommendations must be delivered by an authorized, sworn officer of the court; this is a hard legal and institutional barrier to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	AI systems for document drafting or case summarization would require substantial human oversight by qualified probation officers; the cost of oversight plus inference would exceed the labor savings from a trained professional.
Cost vs. human wage	claude-sonnet-5	2/5	AI could cheaply assist with report drafting and record aggregation, but the investigation and required courtroom testimony still demand paid human officer time, keeping overall cost similar to or only marginally below human-only cost.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably performs prehearing investigations, generates court-admissible presentencing reports, or testifies in court. This remains research-stage and legally infeasible.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product conducts presentencing investigations or testifies in court; this remains firmly a human, in-person, legally mandated function.

Supervise people on community-based sentences, such as electronically monitored home detention, and provide field supervision of probationers by conducting curfew checks or visits to home, work, or school.

CI 0–14 · exposure 8 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Corrections and probation departments are traditionally risk-averse, government-run, and slow to adopt automation in human supervision contexts. Replacement of field supervision by AI faces institutional, legal, and cultural resistance; adoption remains negligible.
Sector adoption velocity	claude-sonnet-5	1/5	Corrections and probation is a slow-adopting, low-digitization public sector with limited AI deployment for in-person supervisory functions.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI could assist probation officers by automating scheduling, flagging at-risk cases via data analysis, or generating compliance reports from case notes, modestly improving productivity. However, the core supervisory and interpersonal work remains human-centered, limiting transformative augmentation.
Augmentation potential	claude-sonnet-5	3/5	AI-enabled electronic monitoring, GPS tracking, and automated alert systems already assist officers by flagging violations and scheduling checks, improving efficiency of the human-led supervision process.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI could assist with scheduling and monitoring electronic data, the core task of in-person curfew checks, home/work/school visits, and field supervision requires human presence and judgment that current AI cannot replicate. Physical supervision and relationship-building with probationers cannot be automated to meet the 50% time-saving threshold.
Task automatability	claude-sonnet-5	1/5	This task requires physical presence for home/work/school visits and in-person judgment about compliance, safety, and behavior that current AI cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Probation officers are licensed/credentialed professionals whose authority to conduct home visits, assess compliance, and make decisions is legally mandated. Many jurisdictions explicitly require a human officer to perform field supervision and sign off on probationer status; liability for failed supervision rests with the human agent.
Adoption barriers	claude-sonnet-5	4/5	Legal and public-safety requirements typically mandate that a certified probation officer conduct supervision and make discretionary judgment calls, creating strong legal/liability barriers to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	AI systems cannot replace the cost of field supervision—visits require human presence and transportation. The infrastructure and human labor for compliance checks and home visits will remain necessary, making AI-based solutions more expensive than existing human supervision.
Cost vs. human wage	claude-sonnet-5	1/5	AI cannot substitute for physical presence, so there is no meaningful AI cost basis to compare against the human officer's wage for this component.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product today reliably performs field supervision visits, in-person compliance checks, or the judgment-intensive assessment of probationer circumstances that this task demands. Electronic monitoring systems exist but do not conduct the human visits and relationship management central to the task.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product conducts physical field visits or in-person curfew checks; electronic monitoring hardware exists but the human supervisory visits remain manual.

Interview probationers and parolees regularly to evaluate their progress in accomplishing goals and maintaining the terms specified in their probation contracts and rehabilitation plans.

CI 0–9 · exposure 5 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Criminal justice agencies typically operate with legacy systems, strong regulatory oversight, and union presence. Adoption of AI for probationer interviews is minimal; the sector remains human-centric and risk-averse due to high stakes and public accountability.
Sector adoption velocity	claude-sonnet-5	1/5	Criminal justice and corrections is a slow-adopting, highly regulated public-sector field with minimal AI deployment for direct offender interaction.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI could assist by transcribing interviews, flagging inconsistencies in statements, suggesting follow-up questions based on prior reports, and auto-generating draft case notes—raising officer productivity. However, the human must remain in full control of the evaluation judgment and legal determination.
Augmentation potential	claude-sonnet-5	2/5	AI could help summarize case notes, flag risk indicators, or prep interview questions, but this offers only modest assistance to the core interpersonal evaluative task.
Task automatability	claude-haiku-4-5-20251001	1/5	While AI could draft summaries of interview notes, this task fundamentally requires real-time interpersonal assessment, judgment of behavioral change, and dynamic rapport-building—elements that current AI systems cannot replicate end-to-end with equal quality. The evaluation of progress against individualized rehabilitation goals demands human clinical judgment that AI cannot yet independently provide.
Task automatability	claude-sonnet-5	1/5	This requires direct human interviewing, reading of nonverbal cues, relationship-building, and judgment calls about risk and compliance that AI cannot perform end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	5/5	This task involves legal authority, risk assessment for public safety, and contractual obligations to the criminal justice system. A licensed probation officer must legally conduct and sign off on evaluations; AI cannot make final determinations about compliance, rehabilitation progress, or recommendations for sanctions without human authorization and accountability.
Adoption barriers	claude-sonnet-5	5/5	Legal and statutory requirements mandate a licensed, authorized probation officer to conduct these evaluations and make supervision decisions, with direct accountability and liability.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI inference for transcription and note summaries is cheap, but the integration, quality oversight, and liability review required would remain substantial. The human labor saved (perhaps 10–20% of interview time) does not offset the cost of maintaining human oversight and error correction for a high-stakes legal task.
Cost vs. human wage	claude-sonnet-5	1/5	Since AI cannot substitute for the interview itself, there is no viable cost comparison—human labor remains mandatory.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed products reliably perform the full interview and evaluation task independently. AI tools exist for note-taking and document analysis, but no mature system currently evaluates probationer progress and rehabilitation compliance in production as a standalone agent.
Technical feasibility today	claude-sonnet-5	1/5	No deployed products conduct autonomous probationer interviews; at most AI assists with note-taking or risk-score analytics after the fact.

Discuss with offenders how such issues as drug and alcohol abuse and anger management problems might have played roles in their criminal behavior.

CI 0–9 · exposure 8 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Correctional systems are traditionally low-tech and risk-averse, with strong professional licensing requirements and labor protections. Adoption of AI for clinical probation functions remains minimal; most jurisdictions still prioritize human-led assessment and treatment.
Sector adoption velocity	claude-sonnet-5	1/5	Corrections and probation agencies are slow-moving, resource-constrained government sectors with minimal AI deployment in direct offender interaction roles.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist by suggesting research-backed topics or tracking conversation themes, but the core therapeutic engagement must remain human-led. The limited potential gain in augmentation reflects the task's heavy reliance on trust, judgment, and human presence.
Augmentation potential	claude-sonnet-5	3/5	AI could help officers prepare discussion guides, summarize case history, or draft follow-up notes, offering moderate but not transformative support for the interpersonal core of the task.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can generate discussion prompts and provide educational content about drug abuse and anger management, it cannot conduct the nuanced, therapeutic dialogue required to help an offender understand their own behavioral triggers. The task requires real-time rapport-building, emotional intelligence, and adaptive responses to individual disclosures that current AI cannot reliably replicate.
Task automatability	claude-sonnet-5	1/5	This requires building trust, reading emotional cues, and adapting a sensitive therapeutic-style conversation with a real person in a supervisory/legal context—far beyond current AI capability to substitute end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Probation and correctional treatment are heavily regulated; in most jurisdictions, a licensed or certified human must conduct these clinical assessments and therapeutic discussions. Additionally, liability and error-cost asymmetry are extremely high—failure to address root causes correctly can directly affect public safety and case outcomes.
Adoption barriers	claude-sonnet-5	5/5	This is a statutorily assigned duty of a sworn/licensed probation officer involving legal accountability, confidentiality, and public safety judgment calls that cannot be delegated to a non-human system.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Even if automation were feasible, the loaded cost of a probation officer is lower than the combination of AI oversight, failure liability, and mandatory human follow-up review needed to ensure compliance and rehabilitation goals are met.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this function, so cost comparison favors the human officer whose judgment, authority, and rapport are required.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs this task. Chatbots exist for mental health support, but correctional settings require licensed professionals because the stakes are legal and rehabilitative; no mainstream system has demonstrated reliable performance in actual probation supervision workflows.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product conducts unsupervised offender counseling conversations about substance abuse or anger management; this remains outside real-world product scope.

Develop liaisons and networks with other parole officers, community agencies, correctional institutions, psychiatric facilities, and aftercare agencies to plan for helping offenders with life adjustments.

CI 0–5 · exposure 0 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	The public sector, particularly corrections, has historically lagged digitization and AI adoption. Probation and parole operate under legacy systems with strong legal and union constraints, limiting experimental AI deployment.
Sector adoption velocity	claude-sonnet-5	1/5	Correctional and social services sectors are slow adopters of AI generally, and interagency liaison work is especially resistant to technological displacement.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist by summarizing institutional contacts, flagging relevant resources, or organizing information about available agencies, but the core task—building trust, negotiating partnerships, and making strategic alliance decisions—remains fundamentally human and cannot be meaningfully augmented by current tools.
Augmentation potential	claude-sonnet-5	3/5	AI tools could help manage contact databases, draft communications, summarize case information, or schedule coordination meetings, offering moderate productivity support to the officer's networking activities.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires sustained relationship-building, strategic judgment about organizational fit, and real-time negotiation across diverse institutional actors—capabilities far beyond current AI. No system can independently develop genuine professional networks or make contextualized partnership decisions that would save ≥50% time at equal quality.
Task automatability	claude-sonnet-5	1/5	This task requires building interpersonal trust, negotiating institutional relationships, and coordinating in-person or relationship-based collaboration across agencies, none of which AI can perform end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Probation and correctional work are heavily regulated; officers must be licensed/certified, carry legal accountability, and represent government authority. Institutions require human-to-human trust and legal responsibility, creating hard barriers to substitution.
Adoption barriers	claude-sonnet-5	4/5	Professional judgment, legal authority over offender supervision, and institutional trust relationships require a credentialed human officer, creating strong organizational and quasi-regulatory barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The human overhead of a probation officer maintaining multi-institutional relationships is substantial, but AI cannot perform this function at all, making cost comparison moot; any AI deployment would require full human oversight, driving costs above human-only execution.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this function, so cost comparison favors the human officer entirely; AI cannot replace the relational capital involved.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs autonomous network development and inter-institutional liaison work. This demands human credibility, accountability, and the ability to navigate complex organizational politics—aspects current AI systems cannot authentically deliver in production.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product performs relationship-building and inter-agency liaison work; this remains fundamentally a human networking and negotiation activity.

Participate in decisions about whether cases should go before courts and which court should hear them.

CI 0–4 · exposure 0 · augmentation 38 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Criminal justice and corrections sectors have been slow to adopt AI for high-stakes decisions; court routing involves legal liability and judicial discretion that organizations resist automating. Production adoption of AI in case-disposition decisions remains minimal.
Sector adoption velocity	claude-sonnet-5	1/5	Criminal justice and corrections is a slow-adopting, heavily regulated public-sector domain with minimal AI deployment for discretionary legal decisions.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist by summarizing case facts or flagging statutory requirements, but the decision itself requires human legal judgment, case context sensitivity, and accountability. Assistance is limited to information retrieval and does not substantially augment the core decision-making task.
Augmentation potential	claude-sonnet-5	3/5	AI can help by summarizing case files, flagging relevant precedents, or organizing jurisdictional criteria, providing moderate assistance while the officer retains full decision authority.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires legal judgment about jurisdictional appropriateness and case readiness—decisions that demand understanding of statute, precedent, and individual circumstances. Current AI systems cannot reliably make or recommend such consequential legal determinations that will be acted upon by courts.
Task automatability	claude-sonnet-5	1/5	This requires discretionary legal judgment, weighing case-specific facts, jurisdictional rules, and human circumstances that current AI cannot reliably navigate end-to-end.rating
Adoption barriers	claude-haiku-4-5-20251001	5/5	Probation officers making case disposition recommendations operate under statutory and judicial oversight; courts ultimately decide jurisdiction and case acceptance. Legal responsibility for routing decisions typically rests with licensed attorneys or officers acting in quasi-judicial capacity, creating hard barriers to full automation.
Adoption barriers	claude-sonnet-5	5/5	Decisions about court jurisdiction and case referral are legally mandated to be made by authorized officers/courts, with significant liability and due-process requirements preventing automation of the decision itself.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of AI-driven errors in routing cases incorrectly (wrong jurisdiction, premature/delayed court involvement) far exceeds the marginal cost of human probation officer involvement. Liability and remediation costs make AI substitution economically unfavorable.
Cost vs. human wage	claude-sonnet-5	2/5	While AI could cheaply provide research or drafting support, the actual decision-making labor is a small, judgment-heavy portion of an officer's role, so cost savings on the core task are minimal.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably makes or participates in court-routing decisions in a production legal context. While AI can assist with legal research, the actual decision to escalate to court and select jurisdiction remains the province of human legal professionals and officers.
Technical feasibility today	claude-sonnet-5	1/5	No deployed products autonomously make jurisdictional or court-referral decisions in probation contexts; this remains a human decision-making process embedded in legal workflows.

Investigate alleged parole violations, using interviews, surveillance, and search and seizure.

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

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Criminal justice and corrections sectors show slow AI adoption overall, with heavy institutional conservatism, union presence, and public resistance to automation of enforcement. Production-level AI deployment for investigation is minimal.
Sector adoption velocity	claude-sonnet-5	1/5	Corrections and law enforcement is a slow-adopting, physically-grounded government sector with minimal AI-driven displacement in field investigative work.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with limited document analysis or flagging patterns in prior violations, but the investigative and interpersonal core of the task—interviews, real-time surveillance decisions, warrant applications—offers narrow scope for augmentation while maintaining human judgment.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with case file review, pattern detection in violation histories, transcription of interviews, and organizing surveillance data, improving efficiency of the surrounding administrative work even though the core physical task remains human-performed.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires judgment-intensive legal and investigative work that depends on human interaction, warrant authorization, and discretionary decision-making. Current AI systems cannot conduct interviews, perform surveillance, execute lawful searches, or make the investigative choices that define this work.
Task automatability	claude-sonnet-5	1/5	This requires physical presence, conducting in-person interviews, executing surveillance operations, and performing legal search and seizure—none of which current AI systems can perform end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong legal and regulatory barriers protect this task: only licensed probation officers or law enforcement can conduct searches and seizures, interviews carry evidentiary weight only when performed by authorized personnel, and warrant requirements mandate human legal judgment.
Adoption barriers	claude-sonnet-5	5/5	Search and seizure requires legal authority vested in a sworn/certified officer, with constitutional protections (Fourth Amendment) and strict chain-of-custody and due process requirements that mandate human authorization and accountability.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	A probation officer's time is moderately specialized labor; the integrated cost of an AI system with sufficient legal and investigative capability, plus necessary human oversight and liability, exceeds what a deployed tool could offer per unit task.
Cost vs. human wage	claude-sonnet-5	1/5	AI cannot substitute for the physical investigative labor involved, so there is no meaningful AI cost comparison—the human officer must be present and legally authorized.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product performs end-to-end parole violation investigation. While AI can assist with document review or data aggregation, the core activities—interviews, surveillance, legal search authority—require human investigators and cannot be reliably automated today.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product conducts physical investigations, surveillance operations, or search and seizure; this remains entirely a human field-based law enforcement function.

Recommend remedial action or initiate court action in response to noncompliance with terms of probation or parole.

CI 0–0 · exposure 0 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Criminal justice and corrections are heavily regulated, union-represented sectors with strong institutional resistance to automation. Adoption of AI for core decision-making in probation enforcement is minimal and faces significant organizational and legal friction.
Sector adoption velocity	claude-sonnet-5	1/5	Criminal justice and corrections agencies are slow-moving, public-sector, low-digitization environments with strong legal and political resistance to automating punitive decisions.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI might assist by surfacing compliance history or flagging violations, but the recommendation and court-action decision remain fundamentally human judgments requiring legal and professional accountability. Modest augmentation potential only.
Augmentation potential	claude-sonnet-5	3/5	AI can help summarize case files, flag compliance patterns, or draft violation reports, assisting officers in preparing recommendations even though final judgment remains human.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires exercising discretionary legal judgment, assessing individual circumstances, and making decisions that trigger formal court proceedings. Current AI cannot independently evaluate noncompliance contexts and recommend court actions—this demands human accountability and judicial authority.
Task automatability	claude-sonnet-5	1/5	This requires exercising discretionary legal and clinical judgment about a person's liberty, weighing case history, risk, and context—something current AI cannot reliably or legitimately perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong legal barriers exist: only licensed probation officers or authorized legal professionals can initiate court action, and case law and statute typically require human judgment and accountability. Liability for incorrect recommendations is high and asymmetric.
Adoption barriers	claude-sonnet-5	5/5	This is a statutorily assigned duty of a sworn/licensed officer with legal authority and accountability; court filings and violation determinations require human authorization and signature.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The loaded cost of a probation officer's judgment and legal accountability far exceeds any inference cost, and there is no commercially viable AI system that could replace this function, making direct cost comparison impossible.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this decision-making function, so cost comparison favors the human by default since AI cannot deliver the output at all.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably recommends court action or initiates judicial proceedings in production. While risk-assessment tools exist, they inform rather than make binding legal recommendations, and initiating court action requires human decision-making and legal authorization.
Technical feasibility today	claude-sonnet-5	1/5	No deployed products make or execute decisions to initiate court action against probationers; this remains a human officer's legal responsibility with no production AI equivalent.

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.