Legislators
11-1031.00Develop, introduce, or enact laws and statutes at the local, tribal, state, or federal level. Includes only workers in elected positions.
Sub-scores
0–100 · band = confidence interval from rater disagreement
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
30 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.
panel mean rating 1.3/5 → substitution pressure 8/100
panel mean rating 1.4/5 → substitution pressure 10/100
panel mean rating 1.6/5 → substitution pressure 15/100
panel mean rating 4.5/5 (barrier strength) → substitution pressure 12/100
panel mean rating 1.3/5 → substitution pressure 8/100
Task breakdown (30 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 knowledge of relevant national and international current events.
49CI 40–59 · exposure 39 · augmentation 88 · click for rater detail
Maintain knowledge of relevant national and international current events.
49| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | While some legislative offices use news tools and AI briefing systems, adoption remains uneven and often partial (supplementing rather than replacing human research); the sector is digitizing but cautiously on core legislative functions. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Political offices and legislative staff increasingly use AI briefing/summarization tools, though adoption is uneven and often supplements rather than replaces staff research. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-powered news aggregation, filtering, and summarization substantially accelerate a legislator's ability to stay informed across multiple domains, allowing them to focus on analysis and decision-making rather than raw information gathering. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI significantly enhances a legislator's ability to stay informed by rapidly summarizing large volumes of news, reports, and international developments, saving substantial research time. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can aggregate and summarize current events from multiple sources efficiently, but legislators must synthesize this information with legal precedent, constituent concerns, and political context—requiring substantial human judgment that AI cannot fully replace at the quality threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can summarize news and current events but 'maintaining knowledge' requires ongoing personal synthesis, judgment about political relevance, and contextualization that current AI cannot fully substitute for at equal quality end-to-end.of a human legislator's needs. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Legislators are accountable for their decisions and must personally understand issues; there is strong institutional and legal expectation that they retain direct knowledge rather than delegate entirely to AI, creating high friction against full automation. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No legal requirement that a human personally consume all information, though political accountability and trust in judgment create some preference for personal engagement with sources. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 5/5 | Automated news monitoring and summarization tools cost far less than the loaded wage of legislative staff or analysts who traditionally perform this work, making AI an order of magnitude cheaper per unit of knowledge maintenance. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI-based news summarization and monitoring tools are far cheaper than paying staff to compile briefings, though some human curation still adds cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | News aggregation, summarization, and alert systems are mature and widely deployed in production; however, the task requires contextual judgment about relevance and significance to legislative priorities, which current systems handle incompletely. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | News aggregation and summarization tools (e.g., AI briefing assistants) are deployed and used by staffers today, but reliability on nuanced political judgment and source vetting remains limited. |
Write, prepare, and deliver statements for the Congressional Record.
38CI 23–54 · exposure 38 · augmentation 75 · click for rater detail
Write, prepare, and deliver statements for the Congressional Record.
38| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislative bodies are notoriously slow-moving and conservative on procedural changes. There is no evidence of AI-driven automation of Congressional Record statement preparation in production; cultural and institutional resistance is substantial. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Legislative offices are professional-services-like environments increasingly using AI for drafting and research, though institutional adoption in government is generally slower than private sector. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by drafting text, organizing policy arguments, and checking facts—useful productivity aids. However, a legislator must ultimately author and validate the substance, limiting the transformative potential of augmentation. |
| Augmentation potential | claude-sonnet-5 | 5/5 | AI is highly useful for drafting, summarizing research, and generating first-pass statements that staff and legislators then edit and finalize. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can draft text and prepare statements, but Congressional Record entries require authentic legislative voice, policy coherence, and legal precision that demand human authorship. Significant human review and rewriting is mandatory, limiting time savings below 50%. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft coherent statements and speeches from talking points, saving significant drafting time, but final delivery, political judgment, and personalization still require human involvement. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | High barriers exist: legislators have personal political accountability for statements entered into the Congressional Record, legal liability for false statements, and democratic legitimacy depends on authentic legislative voice. Regulatory and institutional norms require a human legislator to author and take responsibility. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No legal requirement that a human write these statements, but political accountability, voice authenticity, and risk of embarrassing errors create strong organizational caution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted drafting reduces some writing time, but legislative staff oversight, fact-checking, legal review, and human refinement remain essential and costly. Total cost approaches or exceeds traditional legislator/staff effort. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI drafting of routine statements is far cheaper than staff time per draft, though human review and political vetting remain necessary, moderating full savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI writing tools exist, no deployed system currently handles full end-to-end Congressional Record statement preparation reliably. Products lack understanding of legislative procedure, parliamentary language norms, and the legal weight these statements carry in the formal record. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | LLM drafting tools are widely used by staff to produce first drafts of speeches and statements, but no product autonomously writes and delivers Congressional Record entries end-to-end. |
Oversee expense allowances, ensuring that accounts are balanced at the end of each fiscal year.
38CI 25–51 · exposure 38 · augmentation 63 · click for rater detail
Oversee expense allowances, ensuring that accounts are balanced at the end of each fiscal year.
38| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Legislative bodies are traditionally slow to automate governance and oversight functions; while some use accounting systems, end-to-end automation of expense oversight is rare and adoption remains limited to basic reconciliation aids rather than autonomous oversight. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government and legislative offices tend to be slower adopters of financial automation compared to private-sector finance, due to bureaucratic processes and compliance requirements. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI tools can assist by automatically flagging out-of-balance accounts, categorizing expenses, and preparing reconciliation reports, meaningfully reducing the manual checking workload while the legislator retains final authority and judgment. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-powered expense tracking and anomaly detection tools can significantly speed up reconciliation and flag discrepancies, greatly aiding the legislator or their staff in oversight duties. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can extract and reconcile numerical data from expense records, legislators must exercise discretion on policy compliance, approve exceptions, and make final determinations on allowance allocations—functions requiring human judgment and authority that cannot be fully automated end-to-end today. |
| Task automatability | claude-sonnet-5 | 3/5 | Reconciliation and balancing of expense accounts is a structured financial task that AI/accounting software can largely automate, but oversight and sign-off responsibility remain human.dlings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Legislators typically have personal legal accountability for their own expense accounts, and regulatory/governance frameworks often mandate that elected officials or designated finance committees sign off on expense balances, creating both fiduciary and legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not licensed like an audit, legislative expense oversight often has statutory/ethical accountability requirements meaning a designated official must ultimately certify balances, creating moderate procedural barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted bookkeeping tools are relatively cheap, but the task requires senior personnel (legislators or legislative staff) to perform the oversight function; the all-in cost of human judgment and authority is higher than the marginal AI cost, making replacement uneconomical. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated bookkeeping and reconciliation tools are far cheaper per transaction than dedicated staff time, though legislator-level oversight and final approval still add human cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Accounting software exists to assist with reconciliation and flagging discrepancies, but no deployed product autonomously oversees legislative expense allowances or produces audit-ready final accounts without human review and sign-off by the legislator or finance officer. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Mature accounting and expense-management software (with AI-assisted categorization, anomaly detection, and reconciliation) is widely deployed, though full end-to-end autonomous balancing with judgment calls still requires human review. |
Prepare drafts of amendments, government policies, laws, rules, regulations, budgets, programs and procedures.
34CI 23–46 · exposure 38 · augmentation 63 · click for rater detail
Prepare drafts of amendments, government policies, laws, rules, regulations, budgets, programs and procedures.
34| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislative bodies are among the slowest to adopt AI automation, operate under rigid procedural rules, and view automation of core legislative functions with high public and institutional resistance; adoption remains negligible in practice. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legislative bodies are traditionally slow-moving, hierarchical, and cautious about AI in official processes, though some legislative staff use AI informally for drafting support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist legislators by summarizing prior legislative history, generating initial policy language drafts, identifying precedent, and organizing research—raising productivity on information-gathering and drafting tasks—while the legislator retains final judgment and legal accountability. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI is highly useful for generating first drafts, summarizing precedent, and suggesting amendment language, substantially speeding up the staff work that supports legislators. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can draft routine policy language and summarize existing laws, legislative drafting requires substantive judgment about political feasibility, constitutional implications, stakeholder impact, and alignment with legislative intent—tasks that demand human expertise and accountability. Current AI systems cannot reliably produce end-to-end legislation meeting the 50% time-saving bar without substantial expert review and revision. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft plausible legislative or policy text quickly, but ensuring legal coherence, alignment with intent, political nuance, and jurisdiction-specific formatting requires substantial human revision and negotiation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers protect this task: legislators bear personal legal and electoral accountability for the bills they introduce; constitutional and parliamentary rules often require human authorship and certification; and deep organizational norms favor human authorship as a matter of democratic legitimacy and legal standing. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Actual enactment and formal drafting of legislation typically requires accountable legislative counsel or the legislator's own authorship, with strong institutional and legal norms against non-human drafting of binding law. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While AI drafting tools have low marginal cost, the extensive senior legislative counsel review and revision required to validate legal soundness and policy intent means total cost per production-ready legislative artifact remains comparable to or exceeds a human drafter's loaded wage. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Generating draft text via AI is far cheaper than staff-hours of legal drafting, though oversight and legal vetting still add cost that limits full order-of-magnitude savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably perform full legislative drafting in production; AI tools exist for document generation and research support, but they lack the domain-specific legal and political reasoning needed to generate legally sound, procedurally correct amendments or regulations without expert human oversight and correction. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | LLM tools are already used by staffers to draft bill language and policy memos, but reliability on legal precision and consistency with existing statutes remains inconsistent, requiring expert review. |
Develop expertise in subject matters related to committee assignments.
23CI 11–34 · exposure 17 · augmentation 63 · click for rater detail
Develop expertise in subject matters related to committee assignments.
23| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While legislatures use AI for research support and drafting, actual expertise development remains a personal, deliberate process; adoption of AI as a substitute for this core responsibility has been minimal and faces institutional resistance. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legislative offices are slow, tradition-bound institutions with uneven digitization; AI research tool adoption is emerging but far from deep or standardized. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by rapidly synthesizing research, identifying key stakeholders, and flagging precedent or evidence—useful supports that accelerate a legislator's learning process, though the expertise acquisition itself remains human-driven. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can rapidly summarize legislation, research reports, and expert testimony, substantially speeding up a legislator's or staffer's ability to build subject-matter knowledge. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Developing genuine expertise requires judgment, contextual learning, and sustained engagement with complex policy domains. While AI can retrieve and summarize information, it cannot internalize knowledge or form the strategic understanding that committee expertise demands. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can supply research summaries and background material but the act of developing personal, contextualized expertise—synthesizing constituent priorities, political judgment, and institutional knowledge—remains human-driven and not a fully substitutable output.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislators are personally accountable for their policy knowledge and votes; constituents and the political system expect elected representatives themselves to develop genuine expertise. Legal and democratic norms require human ownership of legislative judgment. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No formal licensing barrier, but political accountability, trust, and the personal nature of policy judgment create real friction against full delegation to AI. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for research and briefing materials are inexpensive, but the core task—developing personal expertise—still requires significant human time investment, and AI only supplements rather than replaces this cost. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI research tools can cut costs of gathering background information significantly, but the ongoing human study, judgment-building, and stakeholder engagement portions still require comparable staff/human time. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI systems can draft background materials and summaries, but no deployed product can independently develop human-level policy expertise or credibly substitute for a legislator's committee preparation and knowledge acquisition. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Products like research assistants and summarization tools exist and are used by staff, but no deployed system independently 'develops expertise' for a legislator in a reliable, production sense. |
Analyze and understand the local and national implications of proposed legislation.
19CI 9–29 · exposure 17 · augmentation 63 · click for rater detail
Analyze and understand the local and national implications of proposed legislation.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislative institutions move slowly and resist automation of core deliberative functions; adoption of AI for implication analysis remains minimal in practice, with most use cases limited to drafting aids and research support rather than decision-driving analysis. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legislative offices are slow-moving, tradition-bound institutions with limited AI deployment beyond occasional staff-level pilot tools. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist legislators by summarizing bills, cross-referencing precedent, identifying stakeholder groups likely affected, and surfacing potential unintended consequences, thereby raising analytical productivity without replacing human judgment on final implications. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully speed up bill summarization, precedent research, and impact drafting, substantially aiding staffers and legislators while they retain final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | While AI can summarize and extract factual content from proposed bills, analyzing implications requires integrating complex political, economic, and social context with nuanced causal reasoning that current systems cannot reliably perform end-to-end. The task demands judgment about downstream effects, stakeholder impacts, and constitutional questions that go beyond pattern matching in available training data. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can summarize bills and surface implications, but genuine political judgment about local constituent impact and national tradeoffs requires contextual, value-laden reasoning current AI cannot reliably perform end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislators have a fiduciary duty to understand legislation before voting; delegation of this core analytical function faces strong institutional and legal barriers. Public accountability, constitutional responsibility, and organizational norms require human lawmakers to personally grapple with implications, creating hard friction against substitution. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Legislators themselves must exercise judgment and are legally/politically accountable for their votes and positions, making full delegation to AI institutionally and legally untenable. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While inference cost is low, the integration overhead, domain expertise required for oversight, and the high cost of errors in legislative analysis mean the all-in cost approaches or exceeds that of a skilled legislative analyst. The human still must verify and contextualize outputs, limiting cost savings. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI drafting/summarization is cheap versus staff analyst time, but the human judgment component still requires paid legislative staff, keeping overall cost comparable. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools can assist by providing bill summaries, identifying related legislation, and flagging potential conflicts, but no deployed product reliably analyzes the full scope of local and national implications at the depth required for legislative decision-making. Systems exist as research prototypes or narrow drafting aids, not as production tools that lawmakers depend on for implication analysis. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some legislative analysis tools and AI summarizers exist in policy offices, but they are used as aids, not as reliable standalone analyzers of legislative implications. |
Evaluate the structure, efficiency, activities, and performance of government agencies.
19CI 13–25 · exposure 20 · augmentation 63 · click for rater detail
Evaluate the structure, efficiency, activities, and performance of government agencies.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislative institutions are traditionally slow to automate core governance functions; adoption of AI for agency evaluation remains minimal and largely limited to exploratory pilots. Political, institutional, and accountability constraints mean this sector lags far behind finance or information services in AI adoption velocity. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legislative bodies are historically slow adopters of AI tools due to political sensitivity, procedural traditions, and public accountability requirements. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist legislators by organizing and summarizing agency data, identifying performance trends, and generating preliminary reports, raising research productivity on parts of the evaluation task. However, the human legislator must remain central to interpreting context, weighing political factors, and making accountability decisions. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can significantly help staff and legislators by synthesizing agency reports, flagging performance metrics, and drafting evaluative summaries, greatly speeding up preparatory work. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can analyze structured data on agency operations, budgets, and performance metrics, the task requires nuanced judgment about governmental effectiveness, political context, and complex institutional dynamics that demand human deliberation. AI can support analysis but cannot independently produce the comprehensive evaluations required for legislative decision-making. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can analyze data and draft reports summarizing agency performance, but the actual evaluative judgment, political weighing of tradeoffs, and stakeholder negotiation central to this task remain beyond current AI capability. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and institutional barriers protect this task: legislators are elected officials with direct accountability to constituents, and evaluations of agency performance are tied to legislative authority and legitimacy. Formal oversight responsibility, legal liability, and democratic accountability create hard barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | This is an inherently governmental/political function often tied to constitutional authority, oversight committees, and elected officials, creating strong institutional and legal barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for data analysis and report generation are relatively cheap, but the loaded cost of a legislator's judgment, stakeholder engagement, and accountability remains far higher than any AI augmentation cost. The human actor cannot be fully replaced, making the all-in cost favorable to humans for this task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply generate data summaries, but the overall task requires expert legislative staff and political oversight, so total cost savings are modest relative to the full task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs this task end-to-end for legislators. Though analytical tools exist for data extraction and report generation, actual legislative evaluation of agency performance requires human expertise, stakeholder consultation, and accountability that current AI systems cannot deliver in production settings. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Analytics and research tools exist to support policy analysts and legislative staff, but no deployed product independently evaluates agency structure and performance as a substitute for legislative judgment. |
Read and review concerns of constituents or the general public and determine if governmental action is necessary.
17CI 5–29 · exposure 17 · augmentation 63 · click for rater detail
Read and review concerns of constituents or the general public and determine if governmental action is necessary.
17| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislative institutions operate under strict democratic and constitutional constraints with minimal demonstrated AI adoption for core representative duties; this task sits at the core of elected office and shows no pattern of displacement even in experimental settings. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legislative offices are generally slow adopters of AI for substantive decision tasks, though some use AI-assisted tools for constituent correspondence triage and summarization. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist legislators by summarizing high-volume constituent mail, clustering concerns by topic, and highlighting patterns or novel issues, improving the legislator's efficiency in reviewing constituent input while the human remains the decision-maker. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by summarizing large volumes of constituent input, detecting trends, and drafting analyses, helping legislators focus attention, while the final judgment remains human. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires nuanced judgment about whether constituent concerns warrant governmental intervention, including interpretation of legitimate grievances, legal standing, political feasibility, and public interest—judgments that demand contextual understanding of constituent intent and alignment with democratic values that current AI cannot reliably replicate end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can help triage and summarize large volumes of constituent messages, but the core judgment—deciding whether governmental action is warranted—requires political judgment, values weighing, and accountability that AI cannot substitute for end-to-end.<br> |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislators are legally and constitutionally required to represent their constituents; there are hard democratic and accountability barriers to delegating the judgment of whether constituent concerns warrant governmental action to unelected AI systems. |
| Adoption barriers | claude-sonnet-5 | 4/5 | This is a core representative duty tied to elected office and democratic accountability; legislators cannot legally or politically delegate the ultimate decision-making function to AI. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Even high-capability AI agents require significant human legislative oversight, fact-checking, and final decision-making, while the task's core value lies in human representative accountability; full cost of AI + human review likely exceeds the cost of human legislators reviewing directly. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cheaply process and summarize large inbound volumes, lowering staff time, but the decision-making component still requires costly human legislative staff and the legislator's own judgment. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can summarize and categorize constituent concerns and flag policy-relevant keywords, no deployed product reliably makes the substantive determination of whether governmental action is necessary—a decision requiring accountability, legal judgment, and constituent representation that remains primarily human. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Constituent-service software and AI triage tools exist to categorize and summarize messages at scale, but no deployed product performs the substantive political judgment of determining if action is necessary. |
Keep abreast of the issues affecting constituents by making personal visits and phone calls, reading local newspapers, and viewing or listening to local broadcasts.
13CI 9–16 · exposure 5 · augmentation 63 · click for rater detail
Keep abreast of the issues affecting constituents by making personal visits and phone calls, reading local newspapers, and viewing or listening to local broadcasts.
13| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislators operate in a domain where personal constituent contact is a core democratic requirement and competitive necessity. There is minimal to no adoption of AI for replacing personal visits and phone calls, as doing so would undermine the legislator's political standing and constituent trust. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legislative offices are slow, tradition-bound institutions with limited AI deployment for constituent engagement beyond basic communications tools. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by summarizing local news trends, flagging emerging constituent concerns from media analysis, and organizing information from broadcasts—helping a legislator prepare for and prioritize personal outreach while remaining in the decision-making role. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can efficiently aggregate news, summarize broadcasts, and flag emerging issues, significantly speeding up how legislators or staff stay informed even though it can't replace personal visits. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task is fundamentally about staying personally connected to constituents through subjective judgment about constituent concerns. While AI can monitor news and broadcasts, the core value—personal visits and calls that build relationships and gauge sentiment—requires human presence and judgment that cannot be meaningfully automated. |
| Task automatability | claude-sonnet-5 | 1/5 | This task depends on personal presence, relationship-building, and direct human interaction with constituents, which AI cannot substitute for legally or practically. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: constituents expect human contact from their elected representative, there is political and reputational risk to outsourcing constituent engagement, and the democratic legitimacy of the role depends on personal accountability that cannot be delegated to AI systems. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Constituents expect and often require direct personal contact with their elected representative; there's no legal requirement but strong political/institutional expectation that a legislator personally engages. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While AI can help aggregate and summarize local information inexpensively, the task's core value—personal constituent engagement—remains labor-intensive. A legislator's time for visits and calls cannot be substituted by automation, making the cost ratio unfavorable for displacement. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply summarize news and broadcasts, but the core value is personal presence and trust-building which requires human labor, keeping overall cost comparable or higher when factoring necessary human involvement. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI systems can assist with monitoring local media and summarizing news, but no deployed product reliably performs the full task of understanding constituent issues through personal contact. Current systems lack the ability to conduct authentic personal visits or phone calls that inform a legislator's understanding of constituent needs. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs personal visits or authentic constituent relationship management; media monitoring tools exist but only cover a small slice of the task. |
Alert constituents of government actions and programs by way of newsletters, personal appearances at town meetings, phone calls, and individual meetings.
13CI 0–25 · exposure 13 · augmentation 50 · click for rater detail
Alert constituents of government actions and programs by way of newsletters, personal appearances at town meetings, phone calls, and individual meetings.
13| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Public sector adoption of AI for core legislative representation tasks is minimal and faces profound political resistance. Legislators remain in traditional, low-digitization patterns of constituent service; pilots and experimental automation in this space are virtually nonexistent. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Government and legislative offices are typically slow adopters of AI for public-facing communication due to political sensitivity, trust concerns, and constituent expectations of personal contact. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by drafting newsletter templates or organizing town-meeting logistics, but the core task of alerting constituents through personal engagement resists meaningful augmentation because authenticity and human relationship are inseparable from the task itself. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can significantly help draft newsletters, summarize constituent concerns, and prepare talking points, meaningfully boosting staff productivity even though the legislator must still personally engage. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task fundamentally requires human judgment about constituent priorities, personalized relationship building, and authentic political representation. While AI could draft newsletters or schedule events, the core task—genuinely alerting constituents through personal appearances, phone calls, and individual meetings—is inherently human-centered and cannot be meaningfully automated to the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can draft newsletters and talking points, but personal appearances, phone calls, and individual meetings require a real legislator's presence and authority, limiting end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Hard barriers exist: legislators have a fiduciary duty to their constituents and must personally engage in representation; delegation to AI for constituent communication would violate public trust norms and likely legal principles of accountability. Constituents have a near-absolute expectation of human contact. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Constituents expect and often legally/politically require direct representation by an elected official; an AI cannot substitute for a legislator's personal accountability and presence. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI systems to automate this task—including legal review, constituent verification, oversight, and reputational risk management—would exceed the cost of a legislator's staff performing outreach, because substitution itself is not legitimate or acceptable. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply generate written content, but the bulk of the task—town halls, phone calls, meetings—still requires paid staff or the legislator's own time, keeping overall costs comparable to human-driven processes. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can reliably perform constituent outreach involving town meetings, phone calls, and individual meetings; doing so would raise immediate legal, ethical, and representational concerns. Current AI lacks the authority, authenticity, and accountability required to act as a legislator in constituent communications. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI drafting tools are used to help write newsletters and communications, but no product substitutes for the personal appearances and constituent interactions central to this task. |
Promote the industries and products of their electoral districts.
10CI 0–20 · exposure 8 · augmentation 38 · click for rater detail
Promote the industries and products of their electoral districts.
10| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This is a foundational legislative function tied to human representation and democratic accountability; no meaningful adoption of AI to replace legislative advocacy exists or would be constitutionally permissible. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Legislative offices are slow, tradition-bound institutions with limited AI integration into core political/advocacy functions, even though some staff use AI tools for drafting support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with research on district industries, data analysis of market trends, or drafting talking points, but the core task of political persuasion and advocacy remains entirely human-dependent and offers limited scope for augmentation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators and their staff research industry data, draft press releases, speeches, and social media content promoting local businesses, improving efficiency in supporting materials. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Promoting industries and products requires political judgment, constituent relationships, negotiation with stakeholders, and strategic decision-making about which sectors to prioritize—all fundamentally human activities. No current AI system can autonomously decide what to advocate for or build the political coalitions necessary to advance these goals. |
| Task automatability | claude-sonnet-5 | 2/5 | Promotion involves relationship-building, public appearances, negotiation, and personal credibility that AI cannot replicate end-to-end, though AI can draft promotional content and speeches.assist with research.assist with speech drafting.and messaging. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislators are elected representatives with legal and constitutional duties to their constituents; only a human with democratic legitimacy can legally and ethically perform the role of promoting district industries. This is a core function that cannot be delegated to non-human agents. |
| Adoption barriers | claude-sonnet-5 | 4/5 | This is an inherently representative political function tied to elected authority and public trust; constituents and businesses expect the actual legislator to advocate, and there's no legal or practical way to delegate this identity-based role to AI. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task requires a human legislator with constituency legitimacy, legal authority, and accountability; AI has no independent ability to advocate politically or command legislative resources, making substitution infeasible on cost grounds. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply generate talking points or marketing copy, but the core value—personal advocacy, credibility, and relationship capital with businesses and constituents—cannot be produced by AI, so cost comparison for the actual task is unfavorable to full automation. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | This task requires representation of constituent interests, political persuasion, and legislative action; no AI system is deployed to autonomously perform legislative promotion or advocacy on behalf of an electoral district. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs district-level economic promotion and advocacy as a substitute for a legislator; this remains a human political and relational function. |
Conduct "head counts" to help predict the outcome of upcoming votes.
7CI 0–14 · exposure 8 · augmentation 38 · click for rater detail
Conduct "head counts" to help predict the outcome of upcoming votes.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This is a core legislative political function performed by human staff and operatives; adoption of AI for legislative head counts is essentially zero because the task is inherently about human political judgment and relationship management. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative bodies are slow to adopt AI for core political functions like whip counts, which rely on personal rapport and confidentiality rather than digitized workflows. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI might assist in organizing public voting records or analyzing past voting patterns, but cannot assist meaningfully in the actual head-counting work, which turns on real-time confidential negotiation and trust-based intelligence gathering that humans must conduct themselves. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help track voting records, sentiment analysis of statements, and organize data on likely votes, providing useful background support to staff conducting the head count. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Head counts require real-time political negotiation, relationship knowledge, informal intelligence gathering, and dynamic judgment about legislator positions that shift based on context—work that depends on human social intelligence and trust that current AI cannot replicate end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | Head counts require personal relationship-building, reading political nuance, private conversations, and persuasion within legislative bodies, which AI cannot conduct autonomously today.》 Some data aggregation could help but the core interpersonal canvassing is not automatable end-to-end.》 , |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Head counts depend on confidential communications with elected officials, informal political networks, and trust-based relationships that cannot be delegated to or replaced by non-human actors; legislators must personally conduct these negotiations. |
| Adoption barriers | claude-sonnet-5 | 4/5 | This task is deeply embedded in political authority, trust, and personal relationships among elected officials; only a legislator or their trusted staff can credibly conduct these counts, creating strong non-regulatory but institutional barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task fundamentally requires embedded human political operatives with relationships and credibility; AI inference cost is irrelevant when the work cannot be performed by AI at all. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this function, so cost comparison favors the human legislator/staff entirely; any AI tool would only supplement, not replace, at added cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs legislative head counts; this task requires access to confidential legislator commitments, ongoing relationship cultivation, and nuanced political interpretation that AI systems do not have in production. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs legislative vote-counting through personal political networking; this remains a human relational and political task with no production AI system replicating it. |
Speak to students to encourage and support the development of future political leaders.
7CI 5–9 · exposure 0 · augmentation 38 · click for rater detail
Speak to students to encourage and support the development of future political leaders.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislative and public-sector engagement with constituents and students remains a low-digitization domain. There is minimal evidence of AI adoption for this task in practice, and political/educational institutions are slow-moving, human-centric sectors where such automation would face cultural and organizational resistance. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Political and civic engagement functions show minimal AI adoption; this remains a deeply human, relationship-based activity with no sectoral push toward automation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist a legislator by drafting talking points, researching student interests, organizing messaging around policy topics, and generating diverse examples to illustrate leadership concepts. These tools can boost preparation and content quality while the legislator maintains personal connection and credibility with students. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could help draft talking points or speech outlines, but it offers little enhancement to the core act of inspiring students through personal presence and authentic engagement. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Speaking to students to encourage political leadership development requires authentic, contextual interpersonal persuasion, real-time responsiveness to audience dynamics, and credibility-building that current AI cannot meaningfully replicate end-to-end. This task is fundamentally about human influence and mentorship. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires a real legislator's personal presence, authority, and lived political experience to inspire students; AI cannot substitute for a human role model or authentic mentorship engagement.atability. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | There are strong barriers to automation: legislators performing this task have legal standing, institutional authority, and personal accountability that cannot be delegated to AI; the task involves direct human contact and trust-building that stakeholders (students, institutions) would resist automating; and there are implicit expectations about authenticity and human judgment. |
| Adoption barriers | claude-sonnet-5 | 4/5 | The task depends on the legislator's personal authority, identity, and social legitimacy as an elected official—an AI cannot be authorized or perceived as a legitimate substitute for this representational role. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI could assist with content preparation and messaging, but the task itself—live speaking and authentic mentorship—requires a human legislator whose wage is far higher than AI inference cost. The irreducible human element makes this economically non-substitutable today. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute service to compare cost against; the human legislator's presence itself is the value delivered, making AI substitution not applicable or cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs this task in production. While AI can draft speeches or generate text, actually engaging students to inspire and guide political leadership development demands human presence, accountability, and adaptive interaction that current systems cannot provide at scale or with authentic impact. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs public speaking or mentorship on behalf of an elected official; this is fundamentally a human presence and credibility task. |
Organize and maintain campaign organizations and fundraisers, in order to raise money for election or re-election.
6CI 5–7 · exposure 0 · augmentation 38 · click for rater detail
Organize and maintain campaign organizations and fundraisers, in order to raise money for election or re-election.
6| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Political campaigns and legislative organizations remain highly personalized, relationship-driven, and resistant to automation. Adoption of AI in campaign management is minimal, with campaigns still relying on traditional grassroots organizing and human-to-human fundraising methods. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Political campaigns use software tools (CRM, donor databases) but organizing and fundraising leadership remains a slow-adopting, relationship-centric domain. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can provide marginal assistance through donor database management, email personalization, and event logistics support, but these are peripheral to the core work of relationship-building and strategic fundraising decision-making that remains human-dependent. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI tools can help with donor list segmentation, drafting fundraising emails, scheduling, and analytics, meaningfully assisting campaign staff even though the human relationship work remains central. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Campaign organization and fundraising require building relationships, political judgment, persuasion, and strategic decision-making that current AI cannot perform end-to-end. While AI can assist with donor research or email drafting, the core activities—cultivating relationships, making strategic decisions, and closing donations—remain fundamentally human tasks. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a relationship-driven, judgment-heavy organizational and persuasion task requiring personal trust and political networking that AI cannot execute end-to-end.5% or so of drafting/logistics could be assisted, but not the core organizing/fundraising work itself. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Campaign finance regulations, FEC compliance requirements, and the legal liability of improper solicitation create material barriers. Additionally, fundraising depends on personal trust and relationships, which require human presence and authorization to be effective and legally compliant. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Campaign finance law, donor relationship norms, and the inherently personal/political nature of fundraising create strong practical and reputational barriers to automation, though not strict licensing requirements. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The human labor involved in campaign management (salaries for campaign managers, coordinators, and fundraisers) is relatively low-cost per unit of work, and AI overhead for donor relationship management, compliance tracking, and strategic decision support would not offer cost savings at comparable quality. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Fundraising success depends on personal donor relationships and trust-building that AI cannot substitute for, so there is no meaningful AI cost comparison for the core task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product autonomously organizes campaigns or raises money at scale. AI tools exist for data analysis and communications support, but no system reliably performs the full campaign management and fundraising function in production. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously runs campaign organizations or fundraising operations; this remains firmly a human relationship-management function. |
Determine campaign strategies for media advertising, positions on issues, and public appearances.
6CI 0–11 · exposure 0 · augmentation 63 · click for rater detail
Determine campaign strategies for media advertising, positions on issues, and public appearances.
6| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Campaigns use AI for media targeting and data analytics, but strategy determination remains firmly in human hands; no sector is automating core campaign strategy decisions, and political accountability structures actively resist delegation to non-human agents. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Political campaigns are adopting AI for micro-tasks like ad targeting or drafting, but adoption of AI in core strategic decision-making remains slow and experimental. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist legislators and campaign teams by analyzing voter data, summarizing media trends, and modeling message testing outcomes, but the human strategist remains firmly in the loop making the final strategic choices and bearing accountability. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully assist with sentiment analysis, message testing, ad copy drafting, and scheduling optimization, substantially boosting staff productivity while the legislator retains final judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Determining campaign strategies requires nuanced political judgment, stakeholder alignment, and accountability for outcomes that current AI cannot handle end-to-end. While AI can assist with data analysis or media research, the core strategic decisions involve values, constituency representation, and political risk that remain fundamentally human responsibilities. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires personal political judgment, values, relationship capital, and strategic authenticity tied to the legislator's own persona; AI cannot end-to-end determine these decisions at equal quality with 50% time savings.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and organizational barriers protect this task: elected officials and campaign leadership bear legal and fiduciary responsibility for campaign decisions; there is no mechanism for AI to assume that accountability, and stakeholders (voters, donors, party leadership) expect human decision-makers. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Elected officials personally bear political and legal accountability for their positions and public statements, and voters expect authentic personal judgment, creating strong barriers to full delegation to AI. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI tools for campaign support (analytics, media monitoring) are far cheaper than the human strategists who set overall campaign direction, but the task itself—strategic determination—remains a high-value human function with no deployed AI replacement. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI-assisted polling analysis or ad copy generation is cheap, the actual strategic decision-making still requires expensive human consultants and the candidate's own judgment, so overall cost savings are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs end-to-end campaign strategy determination; this requires real-time political intelligence, relationship management, and decision authority that organizations intentionally retain with human leadership. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously sets a politician's campaign strategy, issue positions, or appearance schedule; existing tools are only advisory inputs, not decision-makers. |
Seek federal funding for local projects and programs.
4CI 0–9 · exposure 8 · augmentation 50 · click for rater detail
Seek federal funding for local projects and programs.
4| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislators operate in a governance context where their authority and personal relationships are essential. Digitization and AI adoption in this space remains minimal; the task fundamentally depends on human representation and accountability to voters. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative offices are slow to adopt AI for core political advocacy functions, relying on staff and personal networks rather than automated systems for funding requests. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by identifying matching grant opportunities, analyzing eligibility requirements, and drafting portions of funding proposals or briefing documents. However, the core work—building relationships, making the case, and negotiating—remains human-led, so augmentation is useful but secondary. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help draft funding proposals, research precedents, analyze budget data, and prepare talking points, meaningfully assisting staff work that supports the legislator's efforts. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Securing federal funding requires relationship-building, political negotiation, constituency advocacy, and strategic judgment about which grants align with local priorities—capabilities beyond current AI systems. While AI can assist with grant research and proposal drafting, the core task of persuading federal decision-makers demands human political judgment and credibility. |
| Task automatability | claude-sonnet-5 | 2/5 | This involves relationship-building, negotiation, political judgment, and persuasive advocacy with federal agencies and colleagues that current AI cannot execute end-to-end, though drafting support helps a fraction of the work. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Federal funding decisions are the core responsibility and authority of elected legislators. Only humans with legal standing and political legitimacy can authentically represent constituents to federal bodies. This is not merely a procedural task but a legally and constitutionally protected function of elected office. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Legislators are elected officials whose authority to seek and allocate funding is constitutionally and legally tied to their office; no AI can hold this role or legally represent constituents in appropriations processes. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task involves leveraging a legislator's unique political position, constituent mandate, and relationships with federal counterparts—assets that cannot be easily replicated by AI at lower cost. The value delivery depends on human authority and political standing. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | The task's value depends on personal political capital and relationships that AI cannot substitute for, so cost comparison favors humans regardless of AI's low per-token cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can autonomously navigate the political relationships, institutional knowledge, and human-to-human persuasion required to secure federal funding. AI tools exist for grant prospecting and writing support, but none perform the end-to-end task of legislators winning federal resources for their constituencies. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously seeks or secures federal funding through political negotiation; this remains firmly in human-relational territory. |
Hear testimony from constituents, representatives of interest groups, board and commission members, and others with an interest in bills or issues under consideration.
3CI 0–5 · exposure 5 · augmentation 38 · click for rater detail
Hear testimony from constituents, representatives of interest groups, board and commission members, and others with an interest in bills or issues under consideration.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | There is no adoption of AI to replace legislators hearing testimony because the legal and democratic framework explicitly requires human participation. This task is statutorily protected and foundational to legislative authority. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative bodies are slow-moving, tradition-bound institutions with minimal AI adoption for core deliberative functions like hearing testimony, despite some use of AI for research or scheduling support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist by summarizing testimony, organizing witness positions, or flagging key claims, but legislators themselves must listen and judge. The augmentation potential is modest because the core cognitive task—weighing evidence and responding to constituents—remains heavily human-dependent. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by transcribing, summarizing testimony, flagging key points, and helping legislators prepare questions or synthesize input across many witnesses, improving efficiency without replacing the hearing process itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Hearing testimony fundamentally requires human presence, active listening, and real-time judgment about complex arguments. AI cannot serve the legislative function of receiving and weighing constituent input in the way the democratic process requires. |
| Task automatability | claude-sonnet-5 | 1/5 | Hearing testimony is an inherently interpersonal, judgment-laden legislative act requiring physical/social presence, deliberation, and constituent relationship-building that current AI cannot substitute for end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Statutory and constitutional requirements mandate that legislators themselves hear testimony and participate in deliberation. Democratic legitimacy and legal authority require human judgment and presence; substitution would violate legislative procedure and voter expectations. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Legislators are elected officials with constitutional/legal authority to hear testimony and legislate; this role cannot be legally delegated to an AI system, representing a hard structural barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | A legislator must be present to hear testimony as a matter of constitutional and procedural requirement; automation cost is irrelevant since the human cannot be removed. Any AI supplement would only add cost on top of the required legislator. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot substitute for the legislator's role in this task, there is no viable cost comparison—an AI cannot replace the human output being valued (elected judgment and representation). |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can transcribe and summarize testimony after the fact, no deployed system can authentically receive testimony, ask clarifying questions, and exercise the deliberative judgment that legislators must perform during hearings. This remains primarily a human-performed task with only peripheral AI support today. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs this task; there are transcription and summarization tools that assist around the edges but no system that 'hears testimony' as a legislator's representative function. |
Establish personal offices in local districts or states, and manage office staff.
3CI 0–5 · exposure 0 · augmentation 25 · click for rater detail
Establish personal offices in local districts or states, and manage office staff.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No automation or AI adoption is occurring in this task because it is constitutionally and functionally tied to the elected official's personal representation and local presence. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative offices are low-digitization, relationship-driven environments with minimal AI adoption for core administrative/staffing establishment functions. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could marginally assist with scheduling, staff communications, or document management, but the core task of establishing offices and managing relationships requires hands-on human leadership. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI tools can help with scheduling, drafting job postings, or administrative paperwork tied to office setup, but the core establishment and staff management remain human-led with limited AI assistance. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires establishing physical offices and managing human staff relationships, which involves judgment, negotiation, hiring decisions, and interpersonal oversight that cannot be meaningfully automated end-to-end by current AI systems. |
| Task automatability | claude-sonnet-5 | 1/5 | Setting up a physical office presence and hiring/managing human staff involves real-world logistics, personnel management, and interpersonal leadership that AI cannot execute end-to-end.4 |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislators must personally establish and maintain local offices and directly manage staff as part of their elected office responsibilities; legal and constitutional requirements mandate human accountability and presence. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Office establishment involves legal/administrative authority tied to the legislator's role, employment law for hiring staff, and requires human judgment and accountability that cannot be delegated to software. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI has no role in this task; the cost comparison is not applicable since the task fundamentally requires human organizational and staffing decisions. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this task, so cost comparison favors human staff entirely; AI has no direct role to price against. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can perform the core activities of establishing offices and managing staff; these require legal authority, human decision-making, and accountability that must remain with the legislator. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product establishes offices or manages staff; this remains entirely a human administrative and managerial function. |
Appoint nominees to leadership posts, or approve such appointments.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Appoint nominees to leadership posts, or approve such appointments.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | There is zero adoption velocity for automating the appointment decision itself, as legislation and constitutional constraints strictly preserve this power for elected human officials. No sector is moving toward AI-driven legislative appointments. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative appointment processes are governed by law, tradition, and public accountability, showing essentially no movement toward AI-driven decision-making in this specific act. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist with research, candidate background analysis, and vetting preparation, but the core act of appointment—the conferring of authority and responsibility—remains a human legislative function. Augmentation is limited to upstream tasks only. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators research nominee backgrounds, summarize qualifications, or draft materials supporting the decision, though the final appointment judgment remains entirely human. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires making high-stakes personnel decisions that depend on political judgment, stakeholder consensus, and legal/constitutional authority—capabilities that current AI systems cannot perform end-to-end. The decision-making power is vested in human legislators by law and democratic principle; AI cannot serve as a proxy for this discretionary authority. |
| Task automatability | claude-sonnet-5 | 1/5 | Appointing or confirming nominees to leadership posts is an act of political judgment, authority, and legal power vested in an elected official; AI cannot legally or functionally perform this act. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Hard legal and constitutional barriers prevent automation: only human legislators have the legal standing and authority to appoint or approve appointments to leadership positions. Democratic accountability and fiduciary duty require a human decision-maker with personal responsibility. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a constitutionally/statutorily assigned power requiring a duly elected or appointed official's formal vote or signature, representing a hard legal barrier to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | This task has no meaningful cost comparison because the authority to appoint cannot be substituted with AI at any price. The task is not a commodity service but a constitutional or statutory function. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this function at all, so cost comparison is moot; the human legislator's exercise of authority cannot be replaced by cheaper compute. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product performs legislative appointment or confirmation authority; this task is statutorily reserved for human elected officials. AI may assist with research or candidate vetting, but the appointment decision itself cannot be automated or delegated to AI systems. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs official appointment or confirmation actions; this is inherently a human governmental/legal act with no automation product on the market. |
Confer with colleagues to formulate positions and strategies pertaining to pending issues.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Confer with colleagues to formulate positions and strategies pertaining to pending issues.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Legislative bodies are traditionally conservative institutions with high friction around automation. There is no evidence of even pilot adoption of AI-driven legislative strategy formulation; the sector lags significantly in automation. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative bodies are slow-moving, tradition-bound institutions with minimal AI adoption in core deliberative and negotiation functions. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could modestly assist by summarizing colleague positions, drafting background briefings, or analyzing prior voting patterns, but such assistance is peripheral to the core deliberative and political work that remains entirely human-driven. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators prepare briefing materials, summarize positions, or draft talking points ahead of conferring with colleagues, offering moderate productivity support around the core human task. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires negotiation, consensus-building, and nuanced political judgment among human actors with conflicting interests and values. Current AI cannot meaningfully participate in or lead such deliberative processes that depend on accountability, trust, and personal relationships among legislators. |
| Task automatability | claude-sonnet-5 | 1/5 | This is an interpersonal, relationship-driven political negotiation task requiring trust, persuasion, and real-time judgment that current AI cannot perform end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislators have direct legal and fiduciary duty to their constituents; decisions on strategy and positions must be made by accountable elected officials. Democratic legitimacy and accountability requirements create hard barriers preventing AI substitution or autonomous operation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Legislative deliberation and negotiation are constitutionally and institutionally reserved to elected officials, with legal authority, accountability, and political legitimacy requirements that fully block AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task is inherently about human leadership and decision-making that requires accountability and political capital. Any AI assistance would be supplementary to the irreplaceable human negotiation, making it far more expensive to deploy AI than to rely on the legislators themselves. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI product performing this task, so no cost comparison favors AI; the human relational function cannot be replaced by cheaper compute. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs legislative strategy formulation or colleague conferencing. While AI can draft talking points or summarize positions, it cannot authentically participate in the human deliberation, persuasion, and compromise that define this task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product substitutes for legislators conferring with colleagues to build coalitions or strategy; this remains purely a human interpersonal activity. |
Debate the merits of proposals and bill amendments during floor sessions, following the appropriate rules of procedure.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Debate the merits of proposals and bill amendments during floor sessions, following the appropriate rules of procedure.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No sector is adopting AI automation of legislative debate; this task is constitutionally and legally reserved to human elected officials with no pathway to substitution. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislatures are slow-moving, highly traditional institutions with strong procedural and political resistance to any AI substitution in floor proceedings. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with draft language, bill analysis, or precedent research before debate, but the core task of real-time debate participation cannot be augmented without removing the human legislator from the central role. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators research issues, draft talking points, summarize bill language, and prepare rebuttals, but live debate performance itself remains human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Debate requires real-time persuasion, political judgment, constituent representation, and adaptive argumentation in a peer-to-peer adversarial context. Current AI cannot authentically participate as an elected representative accountable to voters. |
| Task automatability | claude-sonnet-5 | 1/5 | Legislative floor debate is a constitutionally and politically embedded human role requiring representation, persuasion, negotiation, and accountability to constituents; AI cannot legally or practically perform this end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Hard legal and constitutional barriers: only elected legislators can debate and vote on the floor under parliamentary rules. The legitimacy and authority of debate rests entirely on human representation and democratic accountability. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Only duly elected legislators may debate and vote on bills under constitutional and procedural rules; this is a hard legal barrier to substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task is inherently tied to a human role with legal authority and electoral accountability; cost comparison is inapplicable since no AI system can legally or legitimately substitute. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no substitute AI system performing this function, so no meaningful cost comparison exists; the human role is irreplaceable in this context. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed system performs legislative floor debate; this requires legal standing, institutional legitimacy, and accountability that cannot be automated. AI can draft talking points but cannot substitute for a legislator's participation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs live legislative debate on behalf of elected officials; this remains entirely outside current AI product deployment. |
Make decisions that balance the perspectives of private citizens, public officials, and party leaders.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Make decisions that balance the perspectives of private citizens, public officials, and party leaders.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | There is no sector-wide adoption of AI for legislative decision-making, nor any trend toward it. The political and legal structures that require human legislators actively prevent and will continue to prevent automation of this core democratic function. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative decision-making is essentially untouched by AI adoption trends; political institutions are slow-moving and resistant to ceding formal decision authority to automated systems. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist legislators by analyzing policy data, summarizing constituent feedback, and modeling outcomes, but these assistive uses are secondary to the core task of making accountable political judgments that remain firmly human-centered. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators synthesize constituent feedback, summarize competing viewpoints, and model policy tradeoffs, usefully informing but not replacing the judgment-based balancing act. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Balancing conflicting political perspectives and making binding policy decisions requires value judgments, stakeholder accountability, and democratic legitimacy that AI cannot replicate. Current AI systems cannot meaningfully replace the political judgment and constituent responsiveness inherent in legislative decision-making. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires legitimate political authority, value judgments, and accountability to constituents that AI cannot exercise; it is fundamentally a human governance function, not an information-processing task susceptible to automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Constitutional law, democratic accountability, and electoral legitimacy create hard barriers: elected officials are legally responsible for legislative decisions and cannot delegate this authority to non-accountable systems. Voters demand human representation and judgment in this role. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Legislators must be elected and are constitutionally/legally required to be the ones making these decisions; this is an unambiguous case of legally mandated human authority. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Even setting aside legal and democratic barriers, the cost of AI oversight and integration to ensure legitimacy and accountability would exceed the alternative of human legislators performing this role, making AI more expensive in practice. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Cost comparison is not meaningful since AI cannot legitimately substitute for the decision-maker; any 'cost savings' would be irrelevant without legal authority to act. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product performs legislative decision-making on behalf of elected officials. While AI can analyze policy trade-offs or summarize constituent views, no system actually makes binding legislative decisions today, as this remains a uniquely human political function. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product makes or is authorized to make legislative decisions on behalf of elected officials; AI is at most used for background research, not decision-making. |
Negotiate with colleagues or members of other political parties in order to reconcile differing interests, and to create policies and agreements.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Negotiate with colleagues or members of other political parties in order to reconcile differing interests, and to create policies and agreements.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No meaningful adoption of AI to conduct political negotiations or replace legislators' role in policy reconciliation is occurring or feasible, as these functions are legally and constitutionally protected. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative negotiation is a highly interpersonal, low-digitization process with essentially no AI agent adoption in production for this specific function. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI might assist legislators by drafting policy language, summarizing opposing viewpoints, or modeling policy impacts, but the actual negotiation, persuasion, and consensus-building remain exclusively human functions requiring political judgment and accountability. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators prepare by summarizing positions, drafting talking points, modeling policy tradeoffs, and analyzing opposition rhetoric, offering moderate support to the human negotiator. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Negotiation with colleagues to reconcile differing political interests and create policies requires genuine deliberation, contextual judgment, trust-building, and understanding of constituency concerns—processes that depend on human credibility, accountability, and democratic legitimacy that AI cannot provide or replace. |
| Task automatability | claude-sonnet-5 | 1/5 | Real-time political negotiation requires trust-building, personal relationships, reading power dynamics, and legal authority to commit—none of which current AI can perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislation and policy-making are explicitly reserved to elected representatives by constitutional and statutory law; AI cannot hold office, sign agreements on behalf of jurisdictions, or negotiate with legal standing. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Only elected, legally authorized legislators can vote, negotiate, and bind their office to agreements; this is a constitutionally and institutionally protected function. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The core task is inherently performed by salaried elected officials whose authority and legitimacy cannot be substituted by AI systems, making any cost comparison inapplicable. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no functioning AI substitute for this task, so no meaningful cost comparison exists; the human is currently the only viable option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed system can autonomously conduct political negotiations, reconcile competing interests, or author binding policy agreements; this task requires human decision-makers with legal authority and electoral accountability. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product conducts inter-party political negotiations on behalf of legislators; this remains entirely research-stage or nonexistent as an application. |
Represent their parties in negotiations with political executives or members of other parties, and when speaking with the media.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Represent their parties in negotiations with political executives or members of other parties, and when speaking with the media.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No adoption is occurring or could occur, as this task is legally reserved to elected officials. The nature of democratic governance creates an absolute barrier to AI substitution. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative and political representation functions show essentially no AI adoption for the core representative/negotiation role itself. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide limited assistance via speech drafting, talking-point generation, or media briefing preparation, but the core task—representing a party with authority and accountability—must remain with the human legislator. Augmentation potential is constrained by the fundamentally human-centric nature of political representation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators prepare talking points, analyze positions, draft statements, and research opposing arguments, meaningfully aiding preparation even though it cannot perform the live representation itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires real-time political judgment, strategic positioning, relationship management, and media communication that depend on human authority and credibility. Current AI systems cannot authentically represent a political party or conduct negotiations with the legitimacy and accountability a legislator provides. |
| Task automatability | claude-sonnet-5 | 1/5 | Political negotiation and public representation require live judgment, trust, authority derived from being an elected human, and real-time relationship management that current AI cannot substitute for end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Extremely strong barriers exist: legislators are elected officials with legal authority to represent their parties, conduct formal negotiations, and make binding commitments. Only a licensed human legislator can legally perform this role; it is constitutionally and legally reserved to elected representatives. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Only an elected legislator has the legal authority and legitimacy to represent a party in negotiations or speak on its behalf; this is fundamentally tied to elected office and political accountability. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Cost comparison is not meaningful here; the task is not automatable at any price because it requires human political authority and accountability that AI cannot assume. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this function, so cost comparison is moot—the human is the only viable performer, making AI effectively far more 'expensive' in the sense of non-substitutable. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed products perform political representation or party negotiation in production. This task requires legal authority vested in an elected human and cannot be delegated to an AI system under any current framework. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs political negotiation or serves as a party's public spokesperson; this remains entirely outside current AI product scope. |
Review bills in committee, and make recommendations about their future.
0CI 0–0 · exposure 0 · augmentation 50 · click for rater detail
Review bills in committee, and make recommendations about their future.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of AI for core legislative functions remains negligible; legislatures use AI only for administrative support (e.g., bill drafting assistance), not for making binding policy recommendations or substituting human deliberation. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislatures are slow-moving, tradition-bound institutions with minimal AI adoption for core deliberative and voting functions. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist with document summarization, identifying precedents, or analyzing bill text, but such tools offer only incremental support for a task whose essence—weighing competing values and political judgment—remains fundamentally human. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist legislators and staff by summarizing bill text, flagging conflicts with existing law, and drafting analysis, improving research efficiency while humans retain judgment and decision-making. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Reviewing bills and making policy recommendations requires deep judgment on complex legal, fiscal, and social implications that demand human values, stakeholder input, and democratic accountability—no AI system today can replicate the deliberative process or produce binding legislative recommendations autonomously. |
| Task automatability | claude-sonnet-5 | 1/5 | Reviewing bills and making recommendations requires political judgment, constituent representation, negotiation, and value-laden decisions that current AI cannot substitute for end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislative review and recommendation is legally and constitutionally reserved to elected representatives; no AI system can legally substitute for or sign off on parliamentary duties, creating absolute adoption barriers. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a constitutionally/statutorily defined role requiring an elected, authorized human representative; no AI system can legally hold office or cast committee votes. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The loaded cost of a legislator's time (salary, staff, institutional overhead) is far lower than any comprehensive AI analysis plus human oversight required for policy decisions of this magnitude. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-only substitute performing this role, so cost comparison favors the human legislator by default since AI cannot deliver the output. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product performs legislative bill review and recommendation-making as a substitute for human legislators; this remains a uniquely human political and legal function with no production automation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs legislative bill review and recommendation as an autonomous replacement for a legislator; this remains entirely a human political function. |
Serve on commissions, investigative panels, study groups, and committees in order to examine specialized areas and recommend action.
0CI 0–0 · exposure 0 · augmentation 50 · click for rater detail
Serve on commissions, investigative panels, study groups, and committees in order to examine specialized areas and recommend action.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No actual adoption of AI serving on commissions or investigative panels is occurring because it is statutorily and constitutionally prohibited. This reflects a laggard sector where automation is legally impossible, not slow adoption. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative bodies are slow-moving, tradition-bound institutions with essentially no adoption of AI agents replacing members' formal committee roles. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist legislators with research summaries, fact-checking, or drafting committee reports, but such assistance is peripheral to the core task of deliberation and recommendation. The human legislator remains fully responsible for the substantive work. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist legislators by summarizing research, drafting background reports, analyzing data, and synthesizing testimony to prepare for committee work, though the human retains the deliberative and decision-making role. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires forming judgments about complex policy issues, weighing competing interests, and building consensus among diverse stakeholders—core human deliberative and political functions that AI cannot perform end-to-end. Current AI systems lack the authority, accountability, and contextual political judgment necessary to serve on official bodies. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires holding elected office, exercising political judgment, negotiating with stakeholders, and taking formal accountability for recommendations—none of which current AI can perform end-to-end.imony required. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task has hard legal and constitutional barriers: legislatures and official commissions are authorized to include only human citizens or appointed officials, often with specific eligibility requirements. Liability, democratic legitimacy, and statutory requirements all require human actors. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Serving on legislative commissions requires being an elected official with legal authority; this is a hard institutional and constitutional barrier that cannot be bypassed by automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI cannot substitute for human service on these bodies; the task itself is fundamentally about human participation and decision-making. Any comparison of cost-per-task is meaningless because no functional automation exists. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this role, so no meaningful cost comparison exists; the human legislator's participation is the deliverable itself. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product can serve on or execute the duties of a commission, panel, or committee in any real legislative or governmental context. AI may assist with research or drafting, but performing the role itself remains entirely out of reach for current systems. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product serves on a commission or panel as a legislator; this is inherently a human institutional role, not a discrete data-processing task. |
Vote on motions, amendments, and decisions on whether or not to report a bill out from committee to the assembly floor.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Vote on motions, amendments, and decisions on whether or not to report a bill out from committee to the assembly floor.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | There is no measurable adoption of AI systems for legislative voting because the practice would be unconstitutional and violate democratic principles; no sector is moving toward this. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Legislative bodies show essentially no movement toward delegating actual votes to AI, though staff may use AI for research and drafting support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI might assist legislators by summarizing bill impacts or analyzing constituency data, but the vote itself remains a human responsibility. Assistance is limited to pre-decision research. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators research bill content, summarize amendments, and analyze impacts prior to voting, improving the information basis for decisions even though the vote itself remains human. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Voting requires human judgment about political values, constituency interests, and legislative priorities—not mere information processing. There is no meaningful way to automate the decision of how a legislator should vote. |
| Task automatability | claude-sonnet-5 | 1/5 | Voting on legislation is a constitutionally vested act of human judgment, representation, and political accountability that cannot be delegated to AI systems today or meaningfully automated. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Voting power is legally vested exclusively in elected legislators. Constitutional and statutory frameworks mandate that votes be cast by the duly elected official themselves, creating an absolute legal barrier to substitution. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Voting is a legally and constitutionally mandated act reserved exclusively for elected officials, representing the hardest possible barrier to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task is performed by elected representatives whose compensation is set by legislative bodies, not market forces. AI cost comparison is not economically meaningful. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this function, so no cost comparison for replacement is applicable; the human cost is irreducible. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system performs legislative voting on behalf of elected officials; doing so would violate democratic representation principles and legal authority vested in individual legislators. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs legislative voting; this is a formal, legally-vested human act with no automation product category. |
Attend receptions, dinners, and conferences to meet people, exchange views and information, and develop working relationships.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Attend receptions, dinners, and conferences to meet people, exchange views and information, and develop working relationships.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This task has not been and will not be adopted for automation in the legislative context, as the legitimacy and effectiveness of the engagement depends entirely on the human legislator's personal participation and authentic relationship-building. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Political and legislative work remains a highly personal, in-person profession with negligible AI displacement of physical networking activities. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could marginally assist by pre-briefing legislators on attendees' backgrounds, interests, or prior interactions, or by summarizing outcomes; however, these are peripheral aids that do not meaningfully amplify the core task of in-person relationship development. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help legislators prepare briefing notes, summarize attendee backgrounds, or draft talking points before such events, offering moderate preparatory assistance without touching the interaction itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task is fundamentally about human social interaction, relationship-building, and informal information exchange at in-person events. Current AI systems cannot physically attend events, read social cues in real-time, or authentically build interpersonal trust—the core value of these activities. |
| Task automatability | claude-sonnet-5 | 1/5 | This is an inherently physical, in-person social and relationship-building task requiring presence, personal rapport, and real-time social judgment that AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legislators are required by their role and constituents to personally attend and engage in these events. There is an implicit—often explicit—requirement that the elected official be present, making substitution by AI effectively impossible due to accountability and representational norms. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Legislative representation requires the elected individual's personal presence and relationships; constituents and colleagues expect direct human interaction, and there is no legal or practical substitute. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The human legislator's presence is the entire deliverable; there is no cost comparison to be made. An AI system cannot substitute for the legislator's physical and social presence at these events. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for physical attendance and relationship-building, so cost comparison is moot—AI cannot deliver the output at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product can perform this task. While AI can help draft talking points or suggest attendees to contact, the actual attendance and relationship-building cannot be automated or delegated to AI systems today. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product attends events, shakes hands, or builds personal political relationships on a legislator's behalf; this remains entirely human territory. |
Encourage and support party candidates for political office.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Encourage and support party candidates for political office.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This task is inherently tied to human elected officials and democratic processes; political sectors have neither incentive nor ability to automate candidate endorsement and support, making adoption velocity near zero. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Political office-holding and campaigning are among the least digitized/automated human activities, with no meaningful AI adoption in this specific function. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI might assist with limited aspects like drafting communication materials or organizing supporter data, but the core function—authentic political judgment and persuasion—remains the legislator's sole responsibility and resists meaningful augmentation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help draft speeches, analyze voter data, or prepare talking points supporting candidates, offering moderate assistance to the legislator's efforts. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires authentic human political judgment, relationship-building, and persuasion—activities involving trust, personal credibility, and face-to-face engagement that AI cannot perform end-to-end. No AI system can meaningfully replace the core political function of endorsing or mobilizing support for candidates. |
| Task automatability | claude-sonnet-5 | 1/5 | This task involves personal political relationship-building, endorsements, and campaigning that require genuine human political credibility and social capital; AI cannot perform this end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal, ethical, and institutional barriers apply: legislators hold elected office with fiduciary duties and political accountability; only humans can make political endorsements or allocate party resources, and constituents expect direct human judgment and representation in candidate support. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Political endorsement and campaigning require an actual elected official's identity, credibility, and legal standing—an AI cannot legally or practically substitute. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task is fundamentally about human political capital, relationships, and decision-making; there is no meaningful AI cost comparison since AI cannot perform the actual work of supporting candidates for office. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this function, so cost comparison favors the human entirely. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product can autonomously encourage or support political candidates in any reliable, scalable way. AI systems cannot make political endorsements, negotiate with party structures, or conduct the relationship-based advocacy this task demands. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product acts as a legislator endorsing or campaigning for candidates; this remains an inherently human political activity. |
Represent their government at local, national, and international meetings and conferences.
0CI 0–0 · exposure 0 · augmentation 38 · click for rater detail
Represent their government at local, national, and international meetings and conferences.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | No adoption of AI substitution for this task is occurring or possible, as it is constitutionally and legally reserved to human elected officials. Adoption velocity is zero by definition. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Government representation functions show essentially no AI adoption trend; this is a highly human, ceremonial and authority-laden domain resistant to automation. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide limited assistance (research summaries, speech drafting, background briefings) but the core act of representation, negotiation, and official participation must remain with the human legislator. Augmentation potential is minimal given the non-negotiable human requirements. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help prepare briefing materials, translate, summarize meeting content, or draft talking points, providing moderate support to the legislator's preparation and follow-up. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Representing government at meetings and conferences requires presence, political judgment, diplomatic negotiation, and accountability that cannot be substituted by AI systems. This task is fundamentally about human judgment, relationship-building, and official authorization. |
| Task automatability | claude-sonnet-5 | 1/5 | Representing a government requires physical/political presence, authority to negotiate and speak on behalf of constituents, and legitimacy that cannot be delegated to AI; no automation is meaningful here. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Hard legal and constitutional barriers exist: only elected or appointed officials can represent government, and delegation of this representative function is explicitly prohibited by law in most jurisdictions. Liability and legitimacy rest entirely on the human legislator. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Legal authority, elected/appointed status, and constitutional requirements mean only the legislator (or designated human official) can represent the government; this is a hard legal and legitimacy barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | A legislator's presence at conferences cannot be cost-effectively replaced by AI at any margin. The task inherently requires a human official, so traditional cost-comparison frameworks do not apply. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this task, so cost comparison favors the human entirely—AI cannot deliver the output at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can legally or functionally represent a government entity at official meetings. This requires a human legislator with formal authority and accountability; no product exists to perform this end-to-end. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No product performs this task; it is inherently a human diplomatic/political representation function, not something deployed AI systems undertake. |
Related occupations — Management
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.