Evidence (5539 claims)
Adoption
5539 claims
Productivity
4793 claims
Governance
4333 claims
Human-AI Collaboration
3326 claims
Labor Markets
2657 claims
Innovation
2510 claims
Org Design
2469 claims
Skills & Training
2017 claims
Inequality
1378 claims
Evidence Matrix
Claim counts by outcome category and direction of finding.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 402 | 112 | 67 | 480 | 1076 |
| Governance & Regulation | 402 | 192 | 122 | 62 | 790 |
| Research Productivity | 249 | 98 | 34 | 311 | 697 |
| Organizational Efficiency | 395 | 95 | 70 | 40 | 603 |
| Technology Adoption Rate | 321 | 126 | 73 | 39 | 564 |
| Firm Productivity | 306 | 39 | 70 | 12 | 432 |
| Output Quality | 256 | 66 | 25 | 28 | 375 |
| AI Safety & Ethics | 116 | 177 | 44 | 24 | 363 |
| Market Structure | 107 | 128 | 85 | 14 | 339 |
| Decision Quality | 177 | 76 | 38 | 20 | 315 |
| Fiscal & Macroeconomic | 89 | 58 | 33 | 22 | 209 |
| Employment Level | 77 | 34 | 80 | 9 | 202 |
| Skill Acquisition | 92 | 33 | 40 | 9 | 174 |
| Innovation Output | 120 | 12 | 23 | 12 | 168 |
| Firm Revenue | 98 | 34 | 22 | — | 154 |
| Consumer Welfare | 73 | 31 | 37 | 7 | 148 |
| Task Allocation | 84 | 16 | 33 | 7 | 140 |
| Inequality Measures | 25 | 77 | 32 | 5 | 139 |
| Regulatory Compliance | 54 | 63 | 13 | 3 | 133 |
| Error Rate | 44 | 51 | 6 | — | 101 |
| Task Completion Time | 88 | 5 | 4 | 3 | 100 |
| Training Effectiveness | 58 | 12 | 12 | 16 | 99 |
| Worker Satisfaction | 47 | 32 | 11 | 7 | 97 |
| Wages & Compensation | 53 | 15 | 20 | 5 | 93 |
| Team Performance | 47 | 12 | 15 | 7 | 82 |
| Automation Exposure | 24 | 22 | 9 | 6 | 62 |
| Job Displacement | 6 | 38 | 13 | — | 57 |
| Hiring & Recruitment | 41 | 4 | 6 | 3 | 54 |
| Developer Productivity | 34 | 4 | 3 | 1 | 42 |
| Social Protection | 22 | 10 | 6 | 2 | 40 |
| Creative Output | 16 | 7 | 5 | 1 | 29 |
| Labor Share of Income | 12 | 5 | 9 | — | 26 |
| Skill Obsolescence | 3 | 20 | 2 | — | 25 |
| Worker Turnover | 10 | 12 | — | 3 | 25 |
Adoption
Remove filter
Core practice 2 — Treat AI outputs as hypotheses: require human sensemaking and validation rather than blind adoption of model outputs.
Prescriptive practice derived from reviewed research and practitioner cases emphasizing human oversight; presented as framework guidance rather than empirically validated intervention.
Core practice 1 — Allocate work by comparative advantage: assign tasks to humans or AI based on relative strengths (e.g., speed, pattern detection, contextual judgement).
Conceptual component of the framework drawn from synthesis of empirical findings in prior human–AI and task allocation literature and practitioner examples; no new empirical testing in the paper.
AI methods have improved molecular property prediction, protein structure modelling, ADME/Tox prediction, NLP-based extraction from literature, virtual screening, and generative chemistry, accelerating early-stage tasks.
Compilation of benchmarking results, method-comparison studies, and applied case studies cited in the paper across these specific application areas.
AI has materially improved efficiency, decision-making, and early-stage productivity in drug discovery, especially in hit discovery, property prediction, and protein modelling.
Synthesis of published benchmarking studies and industry case studies reported in the paper (e.g., improvements in virtual screening throughput, property-prediction benchmarks, and protein-structure prediction results such as those from folding competitions and tool evaluations).
Research agenda priorities include: empirically quantifying the value of digital twins on R&D productivity; studying complementarities between AI tools and tacit sensory knowledge; measuring cultural translation costs; and analyzing market concentration risks from proprietary sensory models.
List of recommended empirical research directions derived from conceptual analysis and gap identification; no primary empirical work conducted within the paper itself.
Seed 2.0 Lite achieved 75.7% success rate with-skill, an increase of +18.9 percentage points over baseline.
Model-specific reported result in the paper: Seed 2.0 Lite with-skill success rate (75.7%) and reported improvement (+18.9pp); reported from the benchmark runs.
GLM-5 Turbo achieved 78.4% success rate with-skill, an increase of +5.4 percentage points over baseline.
Model-specific reported result in the paper: GLM-5 Turbo with-skill success rate (78.4%) and reported improvement (+5.4pp); based on the benchmark evaluation.
Nemotron 120B achieved 78.4% success rate with-skill, an increase of +18.9 percentage points over baseline.
Model-specific reported result in the paper: Nemotron 120B with-skill success rate (78.4%) and reported improvement (+18.9pp); results drawn from the benchmark runs.
MiniMax M2.5 achieved 81.1% success rate with-skill, an increase of +13.5 percentage points over baseline.
Model-specific reported result in the paper: MiniMax M2.5 with-skill success rate (81.1%) and reported improvement (+13.5pp); based on subset of the 185 scenario-runs across the evaluated models.
Results across 5 open-weight model conditions and 185 scenario-runs show consistent skill lift across all models.
Aggregate experimental results reported in the paper: evaluation over 5 model conditions and 185 scenario-runs, with cross-model improvement when SKILL is provided.
Returns to advanced digital skills vary by firm size/type: the wage return in large Chaebol conglomerates is approximately 18.7%, significantly higher than the ~9.5% return in Small and Medium-sized Enterprises (SMEs), indicating a 'skills–scale' complementarity effect.
Heterogeneity analysis within the extended Mincerian wage regression framework using KLIPS micro-data, comparing estimated returns across firm types (Chaebol vs SMEs). (Sample size and exact model specification not provided in the excerpt.)
Workers with only general digital literacy receive a wage premium of approximately 5.8% (after controlling for education, experience, and demographics).
Same empirical framework: extended Mincerian wage equation on KLIPS micro-data with controls for education, experience, and demographic characteristics. (Sample size not specified in the provided excerpt.)
Workers possessing specialized digital skills (e.g., data analysis, programming, automation control) enjoy a significant wage premium of approximately 14.2% after controlling for years of education, work experience, and demographic characteristics.
Empirical estimation using an extended Mincerian wage equation on micro-data from the Korean Labor and Income Panel Study (KLIPS); models control for years of education, work experience, and demographic covariates. (Sample size not specified in the provided excerpt.)
AI-adopting firms increase R&D expenditures following adoption.
Firm financial data showing higher R&D spending for adopters relative to nonadopters in post-adoption periods using the diff-in-diff framework.
Post-adoption patents by AI adopters receive more citations than those of nonadopters.
Difference-in-differences estimates comparing citation counts per patent before and after AI installation versus nonadopters; patent citation data used as the dependent variable.
Firms that adopt AI subsequently increase patenting relative to nonadopters.
Firm-level analysis using a novel AI adoption measure based on timing of AI product installations and a stacked difference-in-differences design exploiting staggered adoption; dependent variable = firm patent counts (patenting rate). (Sample size and exact time period not specified in the provided text.)
Using distributed systems as a principled foundation is a useful approach for creating and evaluating LLM teams.
Primary methodological proposal of the paper; supported by conceptual argument and (per the paper) mappings between distributed-systems concepts and LLM team design (specific experimental validation not detailed in the excerpt).
Large language models (LLMs) are growing increasingly capable.
Statement in the paper's introduction/abstract summarizing the field; based on observed progress in LLM development cited by the authors (no experimental sample size provided in the excerpt).
The work offers a blueprint for converting the ideological potential of AI into implementable, regulator-compatible utilities in pharmaceutical science by synthesizing quantitative measures and practical measures.
Claim about the paper's contribution (blueprint). It is an author claim about the synthesis and guidance provided; the excerpt does not include empirical validation that following the blueprint yields successful implementation.
The paper proposes a systematized framework of integration that emphasizes creating high-impact pilot projects, in-the-wild testing, and ongoing monitoring of models in accordance with FDA, EMA, and EU AI Act guidance.
Described as the paper's proposed framework and recommendations for regulatory-aligned implementation. The excerpt indicates the proposal but does not present validation or empirical testing of the framework.
Grounded in the Resource-Based View (RBV), AI is conceptualized as a strategic intangible resource that can confer a competitive advantage when integrated with complementary capabilities.
Theoretical framing presented in the paper (RBV-based conceptualization); not an empirical finding but an explicit conceptual claim.
Firms with high AI adoption had an average profit growth rate of 9.5%, compared to 5.8% for low adopters.
Reported profit growth rates for high vs. low AI adoption groups from the questionnaire data (N=400); the paper gives the specific averages: 9.5% (high adopters) vs. 5.8% (low adopters).
O artigo discute implicações gerenciais e de políticas públicas para reduzir fricção, acelerar adoção responsável e orientar investimentos em produtividade e inclusão.
Seção de discussão mencionada no resumo abordando encargos gerenciais e políticas públicas; não há avaliação empírica de políticas no resumo.
O artigo entrega instrumentos replicáveis — a escala SCF-30, um checklist de governança mínima de IA e uma matriz 30-60-90 dias — para uso prático.
Afirmação explícita no resumo de que instrumentos replicáveis são disponibilizados; presunção de inclusão dos instrumentos no corpo do artigo.
AI significantly enhances firms' total factor productivity (TFP).
Empirical results from the multidimensional fixed-effects panel model applied to the 2007–2023 sample of agricultural A-share firms; statistical significance reported in the paper.
The model is disciplined using data from the Michigan Survey of Consumers and the Survey of Professional Forecasters, targeting key empirical moments.
Calibration/estimation strategy described in the paper: parameters are chosen to match moments from the Michigan Survey of Consumers and SPF (targeted empirical moments). Specific moments and calibration targets are reported in the paper.
I develop a search-and-matching model with sticky wages and endogenous separations.
Theoretical/model contribution: construction and analysis of a calibrated search-and-matching framework that incorporates wage stickiness and endogenous separation decisions.
Workers and firms face information frictions about the aggregate state of the economy (modeled explicitly).
Assumption and mechanism built into the paper's theoretical framework: a search-and-matching model with information frictions for both sides of the market (model specification).
Households form dispersed, backward-looking expectations about macroeconomic conditions.
Survey evidence from the Michigan Survey of Consumers showing dispersion in individual expectations and patterns consistent with backward-looking (slow/updating) belief formation about macro variables; exact sample sizes and empirical specifications are provided in the paper (not in the summary).
High-quality chatbots (96–100% accurate) improved caseworker accuracy by 27 percentage points.
Experimental result reported in paper: treatment with chatbots at 96–100% aggregate accuracy produced a 27 percentage-point increase in caseworker accuracy compared to control; based on the randomized experiment on the 770-question benchmark.
Caseworker performance significantly improves as chatbot quality improves.
Aggregated results from the randomized experiment show monotonic improvement in caseworker accuracy as the chatbot suggestion accuracy increases; paper states the improvement is statistically significant (specific p-values/statistical tests not provided in the excerpt).
DARE posits that responsible AI deployment requires the simultaneous and integrated development of Digital readiness, Administrative governance, Resilience & ethics, and Economic equity.
Descriptive claim about the framework's components as reported in the abstract (conceptual proposition).
This paper introduces the DARE Framework, a holistic, four-dimensional model for national AI strategy and international cooperation.
Factual description of paper content in abstract — the framework is introduced by the authors (conceptual/model contribution).
ERM is an integrated, strategic framework that aligns risk management with corporate governance, objective setting, and performance management.
Conceptual descriptions and definitions of ERM drawn from existing ERM frameworks and literature reviewed in the article.
The authors curated a set of guidelines called the Incentive-Tuning Framework to aid researchers in designing effective incentive schemes for human–AI decision-making studies.
Authors' contribution described in the paper: development of a framework (framework content and evaluation details not provided in excerpt).
The intelligent scheduling model incorporates legal, contractual, skill-based, and preference-aware constraints to generate equitable and efficient rosters.
Methodological description of constraints encoded in the optimization model for scheduling; experimental validation of resulting rosters reported (conflict reduction and fairness metrics), but specific constraint formulations and datasets are not detailed in the excerpt.
The performance evaluation framework combines structured metrics (task completion, attendance, punctuality) with unstructured feedback (patient surveys, peer reviews) analyzed using natural language processing.
Methodological description in the paper of the performance evaluation module and use of NLP for unstructured feedback analysis; implementation details and dataset sizes not specified in the excerpt.
The proposed AI-driven HRM framework integrates forecasting, optimization, and performance evaluation to enhance workforce planning, staff scheduling, and continuous assessment.
Methodological contribution described in the paper: framework design with three core modules (demand forecasting, intelligent scheduling, performance evaluation); validated via experiments on synthetic and real hospital datasets (dataset sizes not specified in the text).
The Indian government believes that artificial intelligence (AI) will play an important role in India’s continued economic growth, both through its contribution to productivity in the private sector and through smarter and more data-led government.
Reported position in the paper based on review of government statements and policy documents (policy analysis/legal review). No empirical sample size applies; claim is descriptive of government belief.
The positive impact of generative AI on ESG performance is stronger in manufacturing firms, firms in eastern regions, and technology-intensive firms (relative to non-manufacturing, central/western regions, and non-technology-intensive firms).
Heterogeneity/subsample analysis on the panel of Chinese A-share firms (2012–2024) comparing effects across firm types, geographic regions, and technology intensity, showing larger estimated positive effects for manufacturing, eastern-region, and technology-intensive subsamples.
Sustainable innovation partially mediates the relationship between generative AI and corporate ESG performance improvement.
Mediation analysis conducted on the panel dataset (Chinese A-share firms, 2012–2024) indicating a partial mediating role for sustainable innovation measures between generative AI use and ESG performance.
The quality of information disclosure partially mediates the relationship between generative AI and corporate ESG performance improvement.
Mediation analysis (intermediary variable tests) performed on the same panel (Chinese A-share firms, 2012–2024) showing that information-disclosure quality accounts for part of the effect of generative AI on ESG outcomes.
Generative AI can effectively drive improvements in corporate ESG performance.
Empirical analysis using panel data of Chinese A-share listed firms covering 2012–2024; the paper reports an econometric panel-data model showing a positive effect of generative AI adoption/use on measured firm ESG performance.
Convolutional neural networks achieved 95.4% accuracy in identifying ulcers and hemorrhages.
Specific result reported from an included study using convolutional neural networks (accuracy = 95.4%) as cited in the review.
AI tools—ranging from machine learning algorithms in inventory management to natural language processing in customer engagement—are applied in micro‑enterprise contexts.
Descriptive synthesis from included articles reporting specific AI applications (ML for inventory management; NLP for customer engagement) across the reviewed literature.
Global efforts toward establishing shared norms and multilateral cooperation are underway through initiatives led by the United Nations, OECD, UNESCO, and G7.
Qualitative document review identifying initiatives and normative efforts by multilateral organizations (organizations named; specific initiatives referenced qualitatively but not enumerated as a dataset).
Mainstreaming shared input and embracing climate-resilient management approaches are fundamental action items for building institutional resilience.
Paper conclusion lists these recommended action items based on its analysis of governance and sustainability linkages grounded in SDG and global governance literature; the summary does not indicate empirical testing of these recommendations.
Regional peer effects of DT improve firms' resource allocation (RA), which in turn bolsters enterprise resilience (ER).
Mediation/ mechanism analysis on the 2013–2022 Chinese A-share manufacturing panel showing that RA mediates the relationship between regional peer DT and ER.
Industrial peer effects of DT enhance firms' innovation capability (IC), which in turn strengthens enterprise resilience (ER).
Mediation/ mechanism analysis on the same 2013–2022 Chinese A-share manufacturing panel showing that IC mediates the relationship between industrial peer DT and ER.
Digital transformation (DT) exhibits significant industrial and regional peer effects.
Empirical analysis using panel data of Chinese manufacturing enterprises listed on the Shanghai and Shenzhen A-share markets from 2013 to 2022; peer-effect regressions conducted within interlocking directorate networks (IDNs).