Evidence (8807 claims)
Search and filter individual claims pulled from the papers. Looking for a specific finding ("what's the effect on wages?"), you're in the right place. Want to compare whole outcome categories against each other instead? Use the Evidence Explorer.
The board below groups claims two ways: by broad theme (nine paper-level topics) and by outcome category (the 34 claim-level outcomes that the Explorer and Syntheses also use).
Browse by theme
Nine broad, paper-level topics. Click one to filter the claims below.
Adoption
9875 claims
Filter claims →
Productivity
8807 claims
Filtered →
Governance
7870 claims
Filter claims →
Human-AI Collaboration
7560 claims
Filter claims →
Org Design
4892 claims
Filter claims →
Innovation
4781 claims
Filter claims →
Labor Markets
4004 claims
Filter claims →
Skills & Training
3308 claims
Filter claims →
Inequality
2332 claims
Filter claims →
Claims by outcome category
Counts by direction of finding. These are the same 34 outcome categories the Explorer compares and the Syntheses are written for. A linked row has a published synthesis.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 870 | 233 | 116 | 1066 | 2363 |
| Governance & Regulation | 976 | 451 | 218 | 133 | 1809 |
| Organizational Efficiency | 949 | 224 | 144 | 88 | 1416 |
| Technology Adoption Rate | 764 | 287 | 141 | 122 | 1325 |
| Research Productivity | 501 | 152 | 74 | 362 | 1101 |
| Output Quality | 542 | 216 | 69 | 69 | 896 |
| Decision Quality | 387 | 198 | 94 | 54 | 740 |
| Firm Productivity | 513 | 67 | 101 | 27 | 714 |
| AI Safety & Ethics | 249 | 303 | 73 | 36 | 667 |
| Market Structure | 190 | 192 | 134 | 27 | 548 |
| Task Allocation | 243 | 77 | 91 | 36 | 452 |
| Innovation Output | 291 | 33 | 55 | 20 | 401 |
| Skill Acquisition | 206 | 72 | 65 | 21 | 364 |
| Employment Level | 133 | 63 | 115 | 22 | 335 |
| Fiscal & Macroeconomic | 153 | 79 | 52 | 32 | 323 |
| Task Completion Time | 206 | 37 | 12 | 15 | 272 |
| Firm Revenue | 179 | 52 | 29 | 5 | 266 |
| Consumer Welfare | 130 | 76 | 47 | 13 | 266 |
| Inequality Measures | 48 | 137 | 51 | 6 | 242 |
| Worker Satisfaction | 101 | 81 | 25 | 13 | 220 |
| Error Rate | 84 | 110 | 11 | 5 | 210 |
| Wages & Compensation | 98 | 47 | 30 | 10 | 185 |
| Regulatory Compliance | 88 | 73 | 17 | 7 | 185 |
| Automation Exposure | 66 | 64 | 33 | 16 | 182 |
| Team Performance | 105 | 29 | 30 | 11 | 176 |
| Training Effectiveness | 109 | 22 | 14 | 21 | 168 |
| Developer Productivity | 114 | 21 | 14 | 8 | 158 |
| Job Displacement | 12 | 90 | 24 | 1 | 127 |
| Hiring & Recruitment | 57 | 9 | 9 | 5 | 80 |
| Skill Obsolescence | 6 | 56 | 9 | 1 | 72 |
| Social Protection | 43 | 17 | 8 | 2 | 70 |
| Creative Output | 35 | 21 | 9 | 4 | 70 |
| Labor Share of Income | 18 | 21 | 17 | 1 | 57 |
| Worker Turnover | 15 | 16 | — | 4 | 35 |
| Industry | — | — | — | 1 | 1 |
Productivity
Remove filter
Some mechanism-specific estimates are imprecise due to the sample size; confidence intervals for those estimates are wide.
Authors report wide confidence intervals for mechanism decomposition (principal stratification) results based on the randomized sample of 164 students.
There is no consensus in the literature on net job effects — studies diverge on whether AI produces net job gains.
Direct finding from the review: the 17 peer‑reviewed studies produce heterogeneous results on net employment impacts (some positive, some negative, some neutral).
The effects of K_T adoption are heterogeneous across industries, firms, countries, and cohorts — early adopters and capital-rich firms/countries gain most — implying important transition dynamics for political economy.
Cross-country comparisons, industry- and firm-level panel heterogeneity analyses, and case studies demonstrating variation in adoption timing and gains; model simulations emphasizing transition path dependence.
Aggregate productivity (output per worker or per unit of inputs) can rise while labor’s share and employment decline due to substitution toward K_T.
Macro growth-accounting exercises decomposing output growth into contributions from labor, traditional capital, and technological capital; model simulations showing productivity gains coexisting with falling labor shares under substitution elasticities.
AI intensifies market concentration, reinforcing winner-takes-most dynamics through data-driven network effects.
Synthesis of market-structure and industrial-organization studies in the SLR reporting evidence of increased concentration and network/data advantages favoring incumbents.
AI displaces routine occupations.
Synthesis of empirical and modeling studies within the 78-study SLR reporting occupational/task-level substitution effects for routine activities.
Overall, LLM assistance did not produce measurable advantages for human-supervised verification and was associated with reduced detection of major errors, meaning expert human judgment remains indispensable for reliable empirical verification.
Synthesis of experimental findings comparing human-only, AI-assisted, and AI-led conditions; summary concludes no measurable advantages for AI-assistance and reduced major-error detection, and emphasizes continued importance of human expertise.
AI-led teams detected fewer errors across all categories than human or AI-assisted teams.
Reported error-detection comparisons across experimental conditions; summary states AI-led teams detected fewer errors across all categories.
AI-led (autonomous ChatGPT with minimal human oversight) teams achieved only a 37% reproduction rate.
Reported reproduction outcome for AI-led condition in randomized experiment; summary gives 37% reproduction rate for autonomous AI teams.
Verifying results of published social sciences research is expensive, costing hundreds of dollars per study.
Authors' statement in paper background/intro summarizing prior evidence or cost estimates for computational reproducibility efforts; no specific cost study or sample size reported in the provided summary.
In developed regions, DIA–DIT synergy produces negative spatial spillovers on neighbouring areas' green productivity.
Spatial Durbin model results reported in the paper showing negative spillover coefficients for developed regions; summary provides no numeric coefficients or sample size.
The positive effect of DIA–DIT synergy on GP exhibits diminishing marginal returns once the synergy passes a certain threshold.
Threshold models reported in the paper identify a synergy threshold beyond which marginal returns to GP decline; no numeric threshold or sample size provided in the summary.
Dimensional diagnosis identified that 69% of hallucination failures were prompt-induced interpretation errors—these were invisible in aggregate scoring.
Result from the paper's sales-intelligence case study reporting failure-mode breakdown (percentage reported: 69%).
AI has caused a decrease in the labor share of income.
Estimated impacts reported in paper indicate a decline in labor share associated with higher AI exposure; stated as a result of the analysis.
Naively persisting entire conversation histories is token-inefficient and counterproductive because irrelevant context degrades generation quality.
Argumentation in the paper supported by empirical finding that full-history persistence reduced task completion; also conceptual token-efficiency rationale.
Naive full-history persistence actively degrades task completion (by biasing the agent with stale traces) compared to no memory and selective memory.
Empirical comparison reported in the paper showing full-history persistence produced 71% completion vs. 79% for no memory and 96% for selective memory; rationale given that stale reasoning traces bias agents.
There are critical gaps in governance mechanisms that are tuned to the scale of SME deployment of BI and AI.
Conclusion drawn in the narrative review of literature (2020–2025); no specific policy evaluations or sample sizes cited in the excerpt.
SMEs face unequal/fairness issues in access to AI and there are biases in algorithms affecting SME deployment.
Identified as a key gap across the peer‑reviewed literature (2020–2025) in the review; the excerpt provides no quantitative measures or specific studies.
There are critical gaps in data literacy among SME personnel.
Reported as a recurring theme in the reviewed literature (2020–2025) in the narrative review; no numeric prevalence or sample sizes provided in the excerpt.
This structural under‑serving of SMEs by advanced BI and analytics is threatening inclusive economic growth and resiliency.
Argument presented in the review synthesizing literature (2020–2025); no quantified causal estimates or sample sizes provided in the excerpt.
SMEs are systematically under-served by advanced business intelligence (BI) and predictive analytics infrastructure.
Narrative synthesis of peer‑reviewed literature (2020–2025) reported in the review; no specific studies or sample sizes given in the excerpt.
Self-reported cognitive outsourcing predicts lower originality specifically in human-human dyads.
Correlation / regression result from the in-person pilot (N = 62) reporting that self-reported cognitive outsourcing is associated with lower originality in human-human dyads but not in other conditions.
The results caution against using one LLM-generated skill per data-science workflow as a default single-shot prompting strategy.
Authors' interpretation and recommendation based on the null-findings from the ablation and control experiments.
Brown AI’s infrastructure investment crowds out household expenditure, causing the reported consumption cost.
Mechanism described in the paper: modelled exogenous IT investment surge (S3) reallocates resources toward investment and away from household consumption in the CGE results.
These factors (surveillance anxiety, loss of autonomy, deskilling) negatively affect worker well-being and contribute to turnover.
Secondary data literature review of peer-reviewed research and industry evidence published 2022–2026 (method: secondary data review / synthesis). The paper synthesizes prior empirical and theoretical studies but does not report an original sample size.
Automation and algorithmic systems introduce risks of deskilling that affect workers' capabilities.
Secondary data literature review of peer-reviewed research and industry evidence published 2022–2026 (method: secondary data review / synthesis). No primary sample size stated.
Algorithmic management reduces worker autonomy (loss of autonomy) in warehouse settings.
Secondary data literature review of peer-reviewed research and industry evidence published 2022–2026 (method: secondary data review / synthesis). Sample sizes not reported in this paper.
Algorithmic management in automated logistics generates surveillance anxiety among workers.
Secondary data literature review of peer-reviewed research and industry evidence published 2022–2026 (method: secondary data review / synthesis). No sample size given.
Across models spanning 1 billion to 8 billion parameters, controlling output length saves up to 97% of total energy, with the energy dominance of decoding growing stronger at larger model scale.
Empirical measurements across models of sizes 1B–8B parameters within the profiling suite showing energy savings up to 97% when output length is controlled; observed trend of increasing decoding energy dominance with model scale.
Even removing all visual tokens saves at most 10% of total energy for fixed-token models, exposing a fundamental limitation of visual token pruning.
Counterfactual/ablation-style analysis in the profiling study estimating maximum energy savings from eliminating visual tokens in fixed-token model configurations; reported upper bound of ≈10% energy savings.
The paper formalises an AI productivity transmission gap between technical adoption and inclusive productivity realisation.
Formal definition and derivation within the DIAC theoretical framework (analytical/modeling content).
AI does not translate directly from firm-level task efficiency into national productivity; its effect is filtered through complementary intangible investment, skills formation, data governance, competition policy, labor-market mobility, and social insurance.
Analytical DIAC model and accompanying theoretical argumentation in the paper; no empirical sample reported.
AI use can reduce visibility of real skill differences among employees.
Reported findings from performance management and knowledge-work studies indicating that AI-mediated outputs can obscure underlying employee skill variation.
Use of AI can produce over-reliance on AI recommendations, reducing active human judgment and accountability.
Cited empirical observations and prior literature on automation bias and AI-supported decision processes in organizational settings.
AI systems miss contextual information that humans use to make better decisions.
Examples and studies cited from hiring, performance management, healthcare, and knowledge work demonstrating omissions of context by AI tools.
Empirical studies of AI use show recurring problems including mistakes in unusual cases.
Cited recent studies across domains (hiring, performance management, healthcare, knowledge work) reporting AI errors on atypical or edge-case instances.
Human judgment rooted in experience cannot be fully replaced by current AI systems.
Argument based on literature synthesis drawing on cognitive science, neuroscience, and organizational studies; supported by cited recent empirical studies of AI use in hiring, performance management, healthcare, and knowledge work (no single new experiment reported).
The study highlights the limited integration of GenAI in the choice phase of organizational decision-making.
Analysis of task-to-component mappings from the 68 reviewed studies showing relatively fewer GenAI applications mapped to the 'choice' component compared to other components.
Our findings reveal a fragmented application landscape for GenAI in organizational decision-making.
Synthesis of the 68 reviewed publications showing diverse, heterogeneous uses of GenAI across tasks and categories; authors describe the landscape as fragmented.
Existing studies are largely fragmented across industries, organizational contexts, and individual AI applications, with limited systematic evidence synthesizing how AI-aided SIS tools collectively influence organizational performance and sustainable competitive advantage.
Findings from the PRISMA-guided literature search and eligibility assessment that resulted in 22 included studies; thematic analysis highlighted heterogeneity and gaps in the literature.
Despite benefits, challenges persist including data privacy concerns, algorithmic bias, ethical risks, workforce skill gaps, organizational resistance, and high implementation costs.
Recurring themes identified across the 22 studies included in the PRISMA-guided systematic review (Scopus, ScienceDirect, Google Scholar searches, 2017–2026) and summarized via thematic analysis.
The gross tax gap in the U.S is over 600 billion a year.
Statement in paper citing standard U.S. tax-gap estimates (presumably IRS estimates); presented as a factual background statistic in the literature review.
Repository-mining studies measure surface trends but seldom explain the mechanisms beneath them, and the trends themselves prove unstable.
Critical observation by the authors supported by their own GitHub observational analysis showing sensitivity of trends to analysis choices; presented as an interpretive claim in the paper.
Agent-authored pull requests are discussed less than human-authored ones.
Observational analysis of public GitHub activity reported in the paper (no sample size reported in abstract); comparison of discussion volume/length for agent- vs human-authored PRs.
Agent-authored pull requests are reviewed less often than human-authored ones.
Observational analysis of public GitHub activity reported in the paper (no sample size reported in abstract); comparison between agent-authored and human-authored pull requests.
Economic analysis of the information society, digital platforms, and artificial intelligence requires rebuilding the 'hard core' of economic science and abandoning textbook-based learning.
Author's normative/methodological recommendation based on the paper's theoretical critique of existing frameworks and empirical observations about digital sector dynamics.
Market power is shifting to the ownership of the digital assets that underpin markets.
Theoretical and interpretive claim supported by the paper's analysis of digital platforms and asset ownership (no single quantified causal estimate provided).
Digital and even non-digital sectors generate no profit without data, technology, and infrastructure.
Author's theoretical argument and interpretation of contemporary observations (paper's conceptual analysis); not reported as a quantified empirical estimate.
A conceptual model of the AI productivity paradox is proposed to explain underlying causes of efficiency loss and formalize the role of micro-mechanisms in slowing macroeconomic growth.
Theoretical model development drawing on empirical BLS trend analysis and micro-level case evidence; presented as an explanatory framework in the paper.
Key micro-mechanisms underlying the labor productivity paradox under AI are: task expansion, blurring of boundaries between work and non-work time, intensification of multitasking, and accumulation of 'AI debt' by organizations.
Identification and systematization based on theoretical development and analysis of corporate cases and empirical reports.