Evidence (4114 claims)
Adoption
8570 claims
Productivity
7631 claims
Governance
6869 claims
Human-AI Collaboration
6491 claims
Org Design
4175 claims
Innovation
4114 claims
Labor Markets
3566 claims
Skills & Training
2966 claims
Inequality
2066 claims
Evidence Matrix
Claim counts by outcome category and direction of finding.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 758 | 199 | 100 | 900 | 2007 |
| Governance & Regulation | 826 | 400 | 191 | 122 | 1563 |
| Organizational Efficiency | 777 | 193 | 124 | 84 | 1189 |
| Technology Adoption Rate | 635 | 233 | 124 | 97 | 1098 |
| Research Productivity | 422 | 128 | 57 | 336 | 954 |
| Output Quality | 476 | 179 | 59 | 47 | 761 |
| Decision Quality | 328 | 177 | 81 | 47 | 640 |
| Firm Productivity | 435 | 57 | 88 | 20 | 606 |
| AI Safety & Ethics | 218 | 277 | 65 | 33 | 599 |
| Market Structure | 180 | 170 | 123 | 24 | 502 |
| Task Allocation | 213 | 64 | 72 | 33 | 387 |
| Skill Acquisition | 170 | 61 | 61 | 17 | 309 |
| Innovation Output | 203 | 27 | 43 | 18 | 292 |
| Employment Level | 105 | 54 | 107 | 13 | 281 |
| Fiscal & Macroeconomic | 131 | 69 | 43 | 26 | 276 |
| Consumer Welfare | 117 | 63 | 42 | 11 | 233 |
| Firm Revenue | 153 | 48 | 26 | 3 | 230 |
| Task Completion Time | 173 | 31 | 8 | 12 | 225 |
| Inequality Measures | 44 | 122 | 49 | 6 | 221 |
| Worker Satisfaction | 89 | 65 | 22 | 12 | 188 |
| Error Rate | 69 | 92 | 10 | 2 | 173 |
| Regulatory Compliance | 77 | 69 | 14 | 5 | 165 |
| Automation Exposure | 56 | 56 | 26 | 13 | 154 |
| Training Effectiveness | 94 | 21 | 13 | 19 | 149 |
| Wages & Compensation | 77 | 36 | 25 | 6 | 144 |
| Team Performance | 86 | 17 | 27 | 10 | 141 |
| Developer Productivity | 95 | 17 | 14 | 6 | 133 |
| Job Displacement | 12 | 80 | 20 | 1 | 113 |
| Hiring & Recruitment | 52 | 7 | 8 | 3 | 70 |
| Creative Output | 31 | 18 | 8 | 3 | 61 |
| Skill Obsolescence | 5 | 46 | 6 | 1 | 58 |
| Social Protection | 27 | 16 | 8 | 2 | 53 |
| Labor Share of Income | 17 | 19 | 17 | — | 53 |
| Worker Turnover | 11 | 12 | — | 3 | 26 |
| Industry | — | — | — | 1 | 1 |
Innovation
Remove filter
The paper provides empirical evidence that policy tools such as the National AI Innovation and Application Pioneer Zone can help enhance industrial and supply chain security (i.e., SCR).
Analysis was based on the policy of the National AI Innovation and Application Pioneer Zone and authors state their results provide empirical evidence supportive of such policies.
AI's impact on SCR is more significant in enterprises with lower levels of pollution.
Heterogeneity analysis reported by the authors that splits sample by pollution level.
AI's impact on SCR is more significant in private enterprises (versus non-private).
Heterogeneity analysis by ownership type reported in the paper.
AI's impact on SCR is more significant in large-scale enterprises.
Heterogeneity analysis across firm-size categories reported by the authors.
Enterprise agility significantly moderates the AI–SCR relationship: AI's positive effect on SCR is more pronounced in firms with higher agility.
Moderation analysis reported in the paper (moderation models applied to firm-level data).
AI boosts SCR by promoting continuous technological innovation.
Mediation analysis in the paper indicates continuous technological innovation (e.g., R&D/innovation indicators) is a channel through which AI enhances resilience.
AI mainly boosts SCR by improving total factor productivity (TFP).
Mechanism (mediation) analysis reported in the paper using firm-level data; authors identify TFP improvement as a key mediating channel.
The positive effect of AI on SCR holds after multiple robustness checks.
Authors state that the main conclusion remains valid after conducting multiple unspecified robustness checks on the empirical sample (multi-period DID).
AI significantly enhances supply chain resilience (SCR) in manufacturing firms.
Empirical analysis of A-share listed manufacturing companies (2011–2023) using a multi-period difference-in-differences (DID) model; authors report the finding and state it remains after robustness checks.
This study uncovers digital diffusion dynamics and provides theoretical foundations for policymaking.
Paper's concluding statement claiming contributions to understanding diffusion dynamics and policy relevance, based on the analyses (main paths, ERGM, heterogeneity).
In the inter-organizational network, only technological diversity (not proximity) promotes main path formation, indicating knowledge recombination drives micro-level trajectories.
ERGM applied to inter-organizational layer: significant positive coefficient for diversity, non-significant (or not positive) coefficient for proximity; interpretation linking to recombination-driven micro-level diffusion.
ERGM results show that combination opportunities (knowledge recombination potential) consistently promote the formation of main diffusion paths across network layers.
ERGM analysis reporting a positive, significant coefficient for a variable representing combination opportunities or recombination potential.
ERGM results show that technological collaboration value consistently promotes the formation of main diffusion paths across network layers.
Exponential Random Graph Models (ERGM) applied to the multilayer networks; reported positive, significant association between measures of technological collaboration value and presence/formation of main paths.
Geographical technology diffusion networks exhibit a 'core–periphery' structure.
Network analysis of the geographical technology diffusion layer indicating a core–periphery topology across regions.
Inter-organizational diffusion paths center on key universities.
Main path analysis and network mapping of the inter-organizational technology diffusion network showing centrality/positioning of universities in the identified paths.
The patent citation network analysis identifies 14 main paths spanning from core technologies like image recognition to enabling applications.
Main path analysis applied to the patent citation network derived from the patent dataset (2000–2024); result reported as identification of 14 main paths and their topical coverage (e.g., image recognition to applications).
Using patent data of China’s manufacturing digital technologies from 2000–2024, this study constructs a multilayer network comprising patent citation networks, inter-organizational technology diffusion networks, and geographical technology diffusion networks.
Methods reported in the paper: patent dataset covering China's manufacturing digital technologies (years 2000–2024); network construction producing three layers (patent citation, inter-organizational diffusion, geographical diffusion).
Policy implications: strengthening digital infrastructure, human capital, and innovation capacity is important to ensure inclusive productivity gains from the AI revolution in BRICS economies.
Normative recommendation derived from empirical findings that digital infrastructure complements AI-driven TC and EC and that differential AI effects are linked to country-level capacities; recommendation follows from observed divergence across economies.
The study contributes methodologically by providing a comparative, frontier‑based assessment of AI-driven productivity in emerging economies and by distinguishing innovation (frontier-shifting) and diffusion (efficiency) effects of AI.
Two-stage empirical approach combining Malmquist TFP decomposition (frontier analysis) with panel regressions linking TFP components to multiple AI penetration indicators (patents, investment, robot density, digital infrastructure) across BRICS, 2005–2023.
Digital infrastructure is a critical complementary factor influencing both efficiency improvements and frontier‑shifting technological change.
Regression analysis includes digital infrastructure indicators and reports that better digital infrastructure is associated with positive effects on both EC and TC (either directly or via interaction terms with AI indicators). Panel data over BRICS, 2005–2023.
Adoption-oriented AI indicators, including robot density, contribute to efficiency improvements (EC).
Panel regressions linking Efficiency Change (EC) to adoption-oriented indicators (robot density and similar diffusion measures) show positive associations, interpreted as diffusion improving efficiency rather than shifting the frontier.
Innovation-oriented AI activities (AI patents and research investment) are strongly associated with frontier‑shifting technological change (TC).
Second-stage panel regression analysis relating TC to AI penetration indicators (AI patents, AI research investment), using BRICS panel data (2005–2023). Reported statistically significant positive associations between patent/research investment indicators and TC.
China and India exhibit sustained productivity growth over 2005–2023 driven primarily by technological progress.
Malmquist Total Factor Productivity (TFP) index computed for BRICS and decomposed into Efficiency Change (EC) and Technological Change (TC); time series patterns show sustained TFP growth for China and India with TC as the dominant component. Panel covers BRICS economies (Brazil, Russia, India, China, South Africa) for 2005–2023.
Current LLMs are imperfect spatial reasoners, a problem that AADvark addresses by incorporating external constraint solver tools with a specialized visual feedback mechanism.
Diagnosis followed by methodological response: authors argue LLM spatial reasoning is imperfect and describe AADvark's use of external constraint solvers and visual feedback to mitigate this; empirical evidence not provided in this excerpt.
Unlike previous state-of-the-art systems, AADvark captures the dynamic part interactions with one or more degrees-of-freedom.
Design claim about the system's modeling of dynamic part interactions (method/architecture difference); supported by the authors' system design and comparison to prior state-of-the-art as asserted in the paper excerpt.
In this paper we present a prototype of AADvark, an agentic system designed for this task.
Statement of contribution: presentation of a prototype system (methodological contribution described in the paper); evidence would be the prototype and its implementation details (not provided here).
In order for Agent-Aided Design to make a real impact in industrial manufacturing, we need a system that is capable of generating such 3D assemblies.
Normative/argumentative claim by the authors that industrial impact requires capability to generate 3D assemblies with moving parts; no empirical test provided.
In the past year, researchers have started to create agentic systems that can design real-world CAD-style objects in a training-free setting, a new variety of system that we call Agent-Aided Design.
Literature/field observation asserted by the paper (statement of recent research trend); no sample size or empirical count provided in the excerpt.
Digital financial literacy and proper managerial competence are critical for a proper transition of AI outputs into strategic decisions, resulting in a robust governance and regulatory framework for sustainable development (Schrank & Kijkasiwat, 2025, p. 202; Tandilino et al., 2025).
Prescriptive/recommendation claim supported by citations (Schrank & Kijkasiwat, 2025; Tandilino et al., 2025); appears as a policy/managerial implication in the paper rather than an empirically tested result. No sample size or quantitative evidence in the excerpt.
Advanced AI replaces intuition-based decisions with precise and robust data, resulting in a significant increase in the firm's bargaining power during credit negotiations and enabling their access to long term capital (Hamdouni, 2025; Sanga & Aziakpono, 2023).
Assertion supported by citations (Hamdouni, 2025; Sanga & Aziakpono, 2023); framed as a causal pathway (AI -> better data-driven decisions -> increased bargaining power -> improved access to long-term credit). The excerpt does not describe sample size, empirical design, or quantitative estimates.
AI is transforming small business funding by optimizing their internal resources and transitioning the firms from these immediate and short-term loans to long-term capital (Pérez-Campdesuñer et al., 2026; Wu & Liao, 2025).
Claim asserted with citations to Pérez-Campdesuñer et al. (2026) and Wu & Liao (2025); presented as a thematic/finding of the paper (likely based on literature review and RDT framing). No sample size or direct empirical method reported in the excerpt.
GenRec addresses the three listed challenges within a single decoder-only architecture.
Paper claims the proposed GenRec framework (single decoder-only architecture) addresses the three enumerated industrial challenges (method+design claim).
GRPO-SR (Group Relative Policy Optimization with NLL regularization and Hybrid Rewards) aligns generative policy outputs with user satisfaction, provides training stability, and mitigates reward hacking via a dense reward model combined with a relevance gate.
Proposed reinforcement learning method described in the paper (methodological claim about algorithmic design and intended benefits).
An asymmetric linear Token Merger compresses multi-token Semantic IDs in the prompt while preserving full-resolution decoding, reducing input length by ~2X with negligible accuracy loss.
Method description plus reported compression result (~2X reduction) and qualitative statement about accuracy loss in the paper.
Page-wise NTP (next-token prediction) task supervises over an entire interaction page rather than each interacted item individually, providing denser gradient signal and resolving the one-to-many ambiguity of point-wise training.
Proposed training objective described in the paper (methodological claim about training supervision and its intended effects).
In month-long online A/B tests serving production traffic, GenRec achieves 8.7% improvement in transaction count over the existing pipeline.
Reported result from month-long online A/B tests on production traffic (A/B test metric).
In month-long online A/B tests serving production traffic, GenRec achieves 9.5% improvement in click count over the existing pipeline.
Reported result from month-long online A/B tests on production traffic (A/B test metric).
GenRec is deployed on the JD App.
Paper states GenRec was deployed on the JD App (deployment statement).
Continuous learning and diversity of ideas are essential if AI is to play a meaningful role in original scientific discovery.
Normative/conditional claim supported by conceptual reasoning in the article; no empirical evidence or measured sample provided.
AI is likely to fundamentally reshape scientific publication.
Author's argument and discussion of implications for publishing and evaluation; no reported empirical study.
There is a gradual path from AI as a research tool to AI as a scientific collaborator.
Narrative/theoretical progression outlined in the article; conceptual roadmap rather than empirical demonstration.
AI for Science is especially important because it may transform not only the efficiency of research, but also the structure of scientific collaboration, discovery, publishing, and evaluation.
Argumentative/theoretical analysis in the article; forward-looking claim without reported empirical data or experimental sample.
The most important significance of the AI revolution, especially the rise of large language models, lies not simply in automation, but in a fundamental change in how complex information and human know-how are carried, replicated, and shared.
Conceptual argument presented in the article (theoretical/essayistic reasoning); no empirical sample or quantitative study reported.
The conclusions remain robust after substituting different methods for measuring total factor productivity (TFP).
Robustness checks in which alternative TFP measurement methods were used in the panel fixed-effects regressions on the same 2015–2024 sample of Chinese A-share listed firms.
The positive effect of data factor utilization on AI patent output is more pronounced in firms with low total factor productivity (TFP), exhibiting a 'contrarian' catch-up characteristic.
Heterogeneity/interaction analysis in the panel fixed-effects regression dividing firms by TFP level (low vs. high) using the same sample of Chinese A-share listed firms (2015–2024).
The level of data factor utilization has a significant positive impact on AI patent output.
Panel fixed-effects regression applied to a sample of Chinese A-share listed companies in core digital economy industries over 2015–2024; AI patent output used as dependent variable.
For listed firms, AI patents command a robust market-value premium in both countries.
Firm-level analysis linking AI patenting to market valuation for listed firms in both countries (regression or valuation analysis implied by statement).
China surpasses the United States in recent annual AI patent counts.
Time-series patent count comparison using classifier-applied corpora (paper reports that recent annual counts are higher for China than the U.S.).
There is broad convergence in AI patenting intensity and subfield composition between the United States and China.
Comparative analysis of AI patenting intensity and subfield composition across the two patent corpora (US 1976-2023, China 2010-2023) reported in paper.
Applying the classifier to granted U.S. patents (1976-2023) and Chinese patents (2010-2023), we document rapid growth in AI patenting in both countries.
Application of classifier to full corpora of granted U.S. patents (1976-2023) and Chinese patents (2010-2023); time-series counts of AI patents reported.