Evidence (13827 claims)
Adoption
8454 claims
Productivity
7544 claims
Governance
6789 claims
Human-AI Collaboration
6327 claims
Org Design
4126 claims
Innovation
4058 claims
Labor Markets
3520 claims
Skills & Training
2924 claims
Inequality
2057 claims
Evidence Matrix
Claim counts by outcome category and direction of finding.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 749 | 195 | 97 | 889 | 1979 |
| Governance & Regulation | 815 | 391 | 188 | 121 | 1539 |
| Organizational Efficiency | 771 | 189 | 124 | 83 | 1177 |
| Technology Adoption Rate | 624 | 233 | 123 | 96 | 1084 |
| Research Productivity | 410 | 121 | 56 | 331 | 929 |
| Output Quality | 466 | 177 | 59 | 47 | 749 |
| Decision Quality | 320 | 174 | 75 | 42 | 618 |
| Firm Productivity | 435 | 55 | 88 | 20 | 604 |
| AI Safety & Ethics | 214 | 276 | 65 | 33 | 593 |
| Market Structure | 178 | 166 | 122 | 24 | 495 |
| Task Allocation | 206 | 64 | 70 | 31 | 376 |
| Skill Acquisition | 165 | 57 | 60 | 17 | 299 |
| Innovation Output | 201 | 27 | 41 | 18 | 288 |
| Employment Level | 105 | 51 | 107 | 13 | 278 |
| Fiscal & Macroeconomic | 131 | 69 | 43 | 26 | 276 |
| Consumer Welfare | 116 | 63 | 42 | 11 | 232 |
| Firm Revenue | 149 | 46 | 26 | 3 | 224 |
| Inequality Measures | 44 | 122 | 49 | 6 | 221 |
| Task Completion Time | 169 | 29 | 8 | 12 | 219 |
| Worker Satisfaction | 89 | 61 | 20 | 12 | 182 |
| Error Rate | 69 | 91 | 10 | 2 | 172 |
| Regulatory Compliance | 76 | 68 | 14 | 5 | 163 |
| Training Effectiveness | 92 | 19 | 13 | 19 | 145 |
| Wages & Compensation | 77 | 36 | 25 | 6 | 144 |
| Automation Exposure | 51 | 54 | 22 | 12 | 142 |
| Team Performance | 86 | 17 | 27 | 9 | 140 |
| Developer Productivity | 94 | 17 | 14 | 6 | 132 |
| Job Displacement | 12 | 80 | 20 | 1 | 113 |
| Hiring & Recruitment | 51 | 7 | 8 | 3 | 69 |
| Skill Obsolescence | 5 | 45 | 6 | 1 | 57 |
| Creative Output | 31 | 16 | 7 | 2 | 57 |
| Social Protection | 27 | 16 | 8 | 2 | 53 |
| Labor Share of Income | 17 | 17 | 17 | — | 51 |
| Worker Turnover | 11 | 12 | — | 3 | 26 |
| Industry | — | — | — | 1 | 1 |
AI use is concentrated in large firms and knowledge-intensive sectors, reaching 50%–60% (60%–70%, employment-weighted) among very large firms in Information, Professional Services, and Finance.
Stratified descriptive statistics by firm size and industry from the BTOS AI supplement (Nov 2025–Jan 2026); employment-weighted estimates reported; exact sample sizes by stratum not provided in excerpt.
Adoption is expected to reach 22% of firms within six months.
Survey question asking firms about expected near-term adoption (BTOS AI supplement, Nov 2025–Jan 2026), producing a stated expected adoption rate; sample size not given.
Employment-weighted adoption rate was 32% (i.e., 32% of employment is in firms using AI in at least one function).
Employment-weighted descriptive statistic from the BTOS AI supplement covering Nov 2025–Jan 2026 (survey-based weighting by employment; sample size not stated).
During Nov 2025–Jan 2026, 18% of firms used AI in at least one function.
Descriptive statistics from the 2026 AI supplement to the U.S. Census Bureau’s Business Trends and Outlook Survey (BTOS), fielded Nov 2025–Jan 2026; nationally representative firm survey (sample size not stated in excerpt).
Industrial robots influence global value chain length primarily through technological innovation.
Mechanism analysis in the paper linking robot adoption to technological innovation measures and then to GVC length, based on the IFR and 14-subsector panel data; exact innovation indicators and estimation details not provided in the abstract.
Industrial robots influence global value chain length primarily through human capital upgrading.
Mechanism analysis reported in the paper linking robot adoption to changes in human capital (upgrading) and then to changes in GVC length using the same IFR and panel data; specific tests/mediation approaches not detailed in the abstract.
Industrial robots promote participation in global production networks within capital-intensive industries (i.e., they increase global value chain length for capital-intensive sectors).
Subsample or heterogeneous-effects analysis across capital-intensive vs. labor-intensive sub-sectors using the panel of 14 Chinese manufacturing sub-sectors; results reported for capital-intensive industries as positive effect on GVC participation/length.
The application of industrial robots significantly extends the length of global value chains in manufacturing.
Empirical analysis using IFR robot data and panel data on 14 manufacturing sub-sectors; significance reported in paper (panel regression results). Exact model specifications and significance levels not provided in the abstract.
The results generalize existing optimality theorems for fairness-constrained classification and extend them to generalized fairness metrics and partial fairness regimes.
Mathematical generalization and extension of prior theorems to a broader class of fairness metrics and to settings with partial (not full) fairness constraints; proofs provided in the paper.
This result complements existing optimality theorems from the literature which, for specific fairness constraints, posit lower-bound threshold rules only.
Comparative theoretical discussion and extension of prior optimality results (literature comparison plus proofs showing how their characterization extends prior lower-bound-only threshold results).
The Pareto frontier consists of deterministic, group-specific threshold rules applied to individuals' success probability.
Theoretical analysis framing decision making as a multi-objective optimization problem (decision-maker utility vs. group fairness) and deriving the set of Pareto-optimal decision rules for arbitrary utility functions, arbitrary population distributions, and a wide range of group fairness metrics (mathematical proofs/derivations).
Regulatory modernisation, secure national data infrastructure and targeted digital training are essential to enable sustainable innovation in valuation practice.
Policy and practitioner recommendations derived from interview data and thematic analysis; synthesis into prescriptive recommendations.
A majority seems optimistic about [AI's] overall impact.
Paper reports a majority-level positive attitude in surveys about AI's overall impact (no survey details or sample sizes provided in the excerpt).
The framework and results are developed/applied to two instances: AI agent oversight (motivating setting) and marketplace operation (a parallel mechanism-design domain).
Paper includes two instantiated examples/applications illustrating the formal framework: one in AI agent oversight and one in marketplace operation (illustrative case studies within the theoretical paper).
A constructive escape exists: a step-function approval threshold achieves first-best screening for every strictly proper scoring rule, because the agent's binary inflate-or-not choice creates a type-space threshold regardless of the generator's curvature.
Constructive existence proof in the paper showing a step-function approval rule that attains first-best screening; analytical argument based on agent's binary inflate/not strategy.
The principal's optimal oversight necessarily uses a non-affine approval function to screen types.
Analytical result derived from the paper's formal principal-agent model and optimization of the principal's objective (theoretical proof).
The framework is illustrated with applications in income-based social protection programs and humanitarian demining in Colombia, where the tension between screening costs and allocation efficiency is operationally consequential.
Applied examples / case studies presented in the paper (applications to social protection and humanitarian demining contexts).
Efficiency gains from screening grow as the aleatoric uncertainty in the population increases.
Empirical characterization and/or model-based analysis presented in the paper (claims based on theoretical comparative statics and illustrative empirical examples).
In a two-stage allocation framework where a screening stage observes true outcomes for a subset of units before a final allocation under a fixed coverage budget, the optimal strategy screens units at the margin of algorithmic allocation while directly targeting the highest-risk units.
Analytical result derived from the paper's two-stage allocation model (theoretical/mathematical analysis of optimal screening and allocation policy).
Algorithmic targeting is typically cheaper and faster than traditional screening procedures that directly observe the latent vulnerability status through physical verification.
Comparative claim stated in paper introduction; presented as typical advantage of algorithmic targeting (background rationale).
The rise of machine learning has shifted targeted resource allocation in policy and humanitarian settings toward algorithmic targeting based on predicted risk scores.
Descriptive statement in paper introduction; references to the adoption of algorithmic targeting in policy/humanitarian contexts (motivation/background rather than new empirical data).
The contribution of the paper is conceptual and programmatic, presenting a vocabulary, set of candidate primitives, and a research agenda for an agent-aware model of web analytics.
Statement of the paper's scope and contributions (position paper content description; no empirical sample).
The paper proposes five candidate measurement primitives — task chain, actor class, interaction provenance, objective alignment, and signal authenticity — with explicit operational definitions.
Prescriptive contribution of the position paper (proposal based on conceptual synthesis; no empirical validation sample reported).
The paper develops a working taxonomy of crawlers, traditional bots, AI agents, LLM-powered agents, and autonomous agents.
Descriptive contribution of the position paper (conceptual taxonomy developed from literature synthesis; no empirical validation sample reported).
Autonomous AI agents are emerging as a further class of actors layered on top of automated traffic.
Qualitative synthesis/observation of recent developments in AI agents and system design; forward-looking claim in a conceptual paper (no empirical sample presented).
Crawlers and traditional bots already account for a substantial fraction of online interactions.
Assertion grounded in synthesis of existing bot-detection and web-traffic measurement literature cited in the paper (no new empirical sample reported in this position paper excerpt).
Deep learning models (particularly LSTM and Transformer) exhibit stronger tail-risk control than traditional benchmark models.
Empirical risk analysis reported in the paper (tail-risk metrics/comparisons) indicating better tail-risk outcomes for LSTM and Transformer relative to linear and tree-based benchmarks.
Deep learning models (especially LSTM and Transformer) produce more stable WEI scores than traditional benchmarks.
Empirical comparison of WEI (the paper's proposed weighted evaluation index) across model types showing LSTM and Transformer with more stable (less variable/improved) WEI over the evaluation period.
Deep learning models, particularly LSTM and Transformer, deliver superior prediction accuracy compared to traditional benchmarks (linear and tree-based models).
Empirical model comparison using rolling-window forecasts on A-share data (2013–2024) across the listed factors; accuracy metrics reported in the paper (e.g., RMSE or similar) show better performance for deep learning models, with LSTM and Transformer highlighted.
The paper articulates a research agenda for how MASS should be modeled, evaluated and governed.
Stated in the abstract (position paper concludes with an articulated research agenda); evidence is the discussion and proposed agenda sections in the paper.
The importance of each structural prior is demonstrated through formal propositions.
Methodological claim in the abstract that the paper provides formal propositions demonstrating the role/importance of the four priors; evidence contained in proofs/propositions within the paper.
MASS is represented as a class of dynamical systems of information generation, local influence and interaction structure, formulated by four structural priors anchored in social theory: strategic heterogeneity, networked-constrained dependence, co-evolution and distributional instability.
Descriptive claim from the abstract about the formal structure of MASS; supported by the framework and definitions presented in the paper (formal/modeling content).
The paper formalizes a Multi-Agent Social Systems (MASS) framework for how agents interact and influence to generate system-level outcomes.
Direct methodological claim in abstract indicating the authors present a formal framework (MASS) in the paper; evidence consists of the formalization provided in the paper (propositions, definitions).
Agentic AI systems must be modeled with social theory as a structural prior.
Normative / prescriptive claim from the paper's abstract (position paper arguing for this modeling choice; supported by the authors' theoretical arguments and formal framework in the paper).
Emergent dynamics of individuals in a social group have been long studied by social scientists in human contexts.
Historical/contextual claim in the abstract; supported by reference to social-science literature (no sample size; general scholarly consensus).
In multi-agent social settings, system behavior emerges not from individual agents alone, but from the multi-agent interactions over time.
Conceptual claim in the paper's abstract, supported by the paper's argumentation and references to social-science literature on emergent dynamics (formal development likely in main text).
Agentic AI systems are increasingly deployed not in isolation, but inside social environments populated by other agents and humans (e.g., social media platforms, multi-agent LLM pipelines, autonomous robotics fleets).
Statement from the paper's abstract and motivating examples; implied supporting citation/literature review in the paper (no empirical sample size reported in abstract).
Evaluation indicates improved architectural consistency and deployability compared to general-purpose AI code generation workflows, suggesting that constraint-aware retrieval is essential for aligning AI-assisted service development with production software engineering practices.
Paper reports an evaluation comparing the proposed retrieval-augmented scaffolding approach to general-purpose AI code generation workflows and concludes improvements in architectural consistency and deployability; the excerpt does not provide evaluation design details, metrics, or sample size.
By combining template retrieval with structured interaction, the method embeds production-relevant considerations during service scaffolding.
Paper's description of the mechanism by which the proposed approach operates (template retrieval + structured interaction) to incorporate production concerns; presented as a design claim without detailed empirical quantification in the excerpt.
We propose a retrieval-augmented scaffolding approach that combines platform-based code generation with agentic clarification loops to expose and resolve architectural constraint ambiguities.
Methodological contribution described in the paper: a retrieval-augmented scaffolding method combining template retrieval and agentic clarification loops; this is a proposed approach rather than reported empirical proof in the provided text.
AI-assisted development tools enable rapid prototyping of services.
Stated assertion in paper's introduction/abstract that AI-assisted tools speed up prototyping; no quantitative evaluation or sample size given in the provided text.
The C³ Framework provides implementable design patterns and testable propositions intended to help accounting leaders capture productivity gains from human + AI work while preserving accountability, consistency, and alignment with governance expectations in high-stakes reporting contexts.
Conclusions section stating intended practical utility; presented as intended outcomes of applying the proposed framework, not as empirically demonstrated results in this paper.
The paper proposes a role taxonomy that clarifies review responsibility, escalation thresholds, and evidence retention for human–AI collaboration in accounting.
Results section proposing a role taxonomy as part of the C³ Framework; presented as a design artifact derived from synthesis of research and guidance.
The framework specifies five mandatory control points for high-judgment use cases: source grounding and traceability, independent verification and tie-out, contradiction testing, escalation and approval, and audit-trail logging.
Results section listing five control points as mandatory design elements for high-judgment accounting use cases; conceptual recommendation from synthesis.
The paper develops the C³ Framework—Complementarity, Controls, and Competencies—which maps accounting tasks by task structure and judgment/materiality to recommend collaboration modes.
Results section: conceptual framework developed by the authors based on synthesized literature and guidance; no reported empirical validation in the abstract.
AI accelerates drafting, summarization, and pattern detection in accounting while professionals remain accountable for judgment, materiality, and defensibility in financial reporting and analysis.
Statement in paper summarizing literature and practitioner guidance (2023–2025); conceptual synthesis rather than new empirical data.
AI tools can serve as valuable aids in task splitting, provided there is human oversight to filter out irrelevant tasks.
Paper's conclusion synthesizing experimental results and participant feedback, recommending human-in-the-loop oversight when using AI for task-splitting.
Participants favored a hybrid approach, combining AI tools with conventional methods to maintain high accuracy in planning.
Participant preferences and qualitative feedback reported from the controlled experiment indicating preference for combining AI assistance with human methods; sample size not provided.
AI-assisted approaches can help ensure no important tasks are overlooked during task-splitting.
Reported finding from the experiment indicating AI assistance reduced omissions in task lists (paper statement based on experiment and participant observations); sample size not stated.
AI-assisted approaches can generate more granular task lists than traditional methods.
Experimental comparison reported in the paper showing AI-generated task lists were more granular (based on task lists produced during the controlled experiment); sample size not provided in summary.