Evidence (600 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
21267 claims
Filter claims →
Productivity
17978 claims
Filter claims →
Governance
17038 claims
Filter claims →
Human-AI Collaboration
16914 claims
Filter claims →
Org Design
11104 claims
Filter claims →
Innovation
11087 claims
Filter claims →
Labor Markets
6711 claims
Filter claims →
Skills & Training
5616 claims
Filter claims →
Inequality
4343 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 | 1880 | 496 | 296 | 1854 | 4721 |
| Organizational Efficiency | 2906 | 665 | 438 | 180 | 4210 |
| Governance & Regulation | 2162 | 929 | 480 | 247 | 3866 |
| Technology Adoption Rate | 1533 | 545 | 278 | 210 | 2593 |
| Decision Quality | 1391 | 534 | 321 | 173 | 2429 |
| Output Quality | 1298 | 472 | 231 | 145 | 2153 |
| AI Safety & Ethics | 682 | 821 | 230 | 90 | 1837 |
| Research Productivity | 855 | 253 | 121 | 425 | 1675 |
| Firm Productivity | 1105 | 171 | 175 | 73 | 1531 |
| Task Allocation | 735 | 229 | 361 | 99 | 1433 |
| Market Structure | 457 | 461 | 251 | 47 | 1222 |
| Innovation Output | 673 | 94 | 108 | 36 | 913 |
| Task Completion Time | 499 | 118 | 43 | 38 | 702 |
| Firm Revenue | 458 | 130 | 61 | 26 | 677 |
| Skill Acquisition | 381 | 122 | 113 | 34 | 650 |
| Consumer Welfare | 316 | 176 | 115 | 39 | 648 |
| Employment Level | 223 | 143 | 177 | 53 | 600 |
| Error Rate | 246 | 282 | 44 | 19 | 594 |
| Fiscal & Macroeconomic | 283 | 142 | 78 | 52 | 562 |
| Inequality Measures | 103 | 329 | 106 | 13 | 552 |
| Worker Satisfaction | 225 | 185 | 63 | 30 | 503 |
| Automation Exposure | 158 | 155 | 72 | 37 | 426 |
| Regulatory Compliance | 186 | 126 | 35 | 14 | 362 |
| Team Performance | 193 | 56 | 51 | 24 | 326 |
| Developer Productivity | 224 | 58 | 27 | 13 | 323 |
| Wages & Compensation | 148 | 108 | 50 | 17 | 323 |
| Training Effectiveness | 218 | 44 | 21 | 27 | 313 |
| Job Displacement | 23 | 159 | 53 | 5 | 240 |
| Hiring & Recruitment | 109 | 61 | 32 | 11 | 215 |
| Skill Obsolescence | 16 | 107 | 26 | 6 | 155 |
| Creative Output | 71 | 44 | 28 | 6 | 150 |
| Social Protection | 58 | 31 | 12 | 3 | 104 |
| Labor Share of Income | 29 | 43 | 25 | 2 | 99 |
| Worker Turnover | 45 | 29 | 6 | 4 | 84 |
| Industry | — | — | — | 1 | 1 |
AI-driven automation may displace certain categories of labor while increasing demand for other categories, potentially altering wage structures, income distribution, and the relative shares of labor and capital in national income.
Theoretical discussion of labor substitution, labor complementarity, and distributional effects; no labor-market sample or measured effect is reported.
AI-driven automation may displace certain categories of labor while increasing demand for other categories, potentially altering wage structures, income distribution, and the relative shares of labor and capital in national income.
Theoretical discussion of labor substitution, labor complementarity, and distributional effects; no labor-market sample or measured effect is reported.
AI creates new roles while transforming or displacing tasks that are routine or highly automatable.
Synthesis of labor-market studies using administrative, survey, and matched employer–employee data, alongside task-based and occupational analyses.
Short-run disruption includes job churn and wage compression for affected groups, while long-run outcomes depend on reskilling, capital re-allocation, and institutions.
Asserted in the supplied example contribution; no longitudinal employment, wage, or reskilling evidence is provided.
Net employment effects are modest short-run losses, with potential long-run gains if complementary skill investment and policy support occur.
Asserted in the supplied example contribution; the text provides no employment panel, identification strategy results, or quantified estimates.
Occupations made automatable by Wave 1 show a slight employment decline of about 1%, while occupations first made automatable at Wave 2 or later show flat or rising employment.
Employment-weighted US OEWS changes for 2023–24 and 2024–25, grouped by the wave in which occupations first meet all nine AI capability requirements.
The paper argues that age influences employment-transition outcomes indirectly through digital literacy and access to training opportunities rather than acting as a deterministic factor.
The claim is stated in the abstract and conceptual framework; the supplied text does not report age-stratified estimates or a formal moderation analysis.
The paper argues that adaptive capacity is the core mediator of employment divergence following AI-related employment shocks.
This is the study's stated interpretive conclusion, derived from questionnaire evidence and grounded-theory-informed coding involving frontline workers and managers; no formal mediation analysis is reported.
Frontline workers expressed both fear of unemployment and expectations that AI could create new employment opportunities.
Q19 coding found 16 responses (10.70%) expressing fear of unemployment and 16 responses (10.70%) identifying new employment opportunities.
Automation of audit, risk-scoring, and tax-processing tasks is expected to reconfigure public-sector labor demand toward data-science and governance roles.
Economic interpretation of the likely labor-market effects of automating tax-administration tasks; no employment dataset or causal estimate is reported.
AI contributes to employment polarization by increasing the relative demand for high-skilled and low-skilled labor while reducing demand for medium-skilled labor.
The paper provides a conceptual task-based explanation: AI is relatively advantaged in routine cognitive and physical tasks, while non-routine abstract and manual tasks remain more dependent on human labor; it cites prior empirical research.
The World Economic Forum projects a net global gain of 78 million jobs by 2030, alongside 92 million job losses in more automatable categories, with software- and AI-related roles among the fastest-growing.
World Economic Forum Future of Jobs projection cited by the report; this is a forecast rather than an observed causal estimate.
Employment among developers aged 22–25 fell nearly 20% from its late-2022 peak between 2021 and mid-2025, while employment among more experienced developers grew by approximately 6–12%.
Payroll-based labor-market research covering 2021–2025; the paper explicitly characterizes these labor-market findings as correlational rather than causal.
The net employment effect of AI adoption depends on the balance between task displacement and the creation of new tasks.
Conceptual economic interpretation in the paper's implications section; no direct employment estimate is reported.
AI adoption in banking is associated with skill polarization: demand for routine, low-skill tasks declines while demand increases for high-skill technical roles such as data analysts, AI engineers, and cybersecurity specialists.
Qualitative synthesis of the literature's reported labor impacts; the supplied text does not report a causal identification strategy or quantitative labor-market estimates.
The three organizations studied—Testworks, Kakao, and Microsoft—use AI to widen labor-market access, but none integrates older adults and persons with disabilities within a single sociotechnical system.
This finding comes from the paper's multiple-case cross-case analysis of three theoretically distinct organizations.
The employment effect of AI for persons with disabilities changes from negative to positive beyond critical levels of AI adoption.
The paper summarizes Abid et al. (2024), which examined linear and nonlinear effects; the underlying sample size and numerical thresholds are not reported in the supplied text.
Net employment effects of AI vary by sector and occupation rather than following a uniform positive or negative pattern.
Cross-sector descriptive comparison of employment trends and occupational AI exposure; the study explicitly does not make causal claims.
The study operationalizes organizational restructuring through the share of employment in middle-management occupations, specifically ISCO-08 groups 12–13.
Research design described in the introduction; the dependent variable is an employment-structure proxy rather than a direct firm-level measure of restructuring.
The strongest predictors of AI job growth in US counties were the share of STEM degrees, local labor-market tightness, and patenting activity, while manufacturing intensity was negatively associated with AI job growth.
Andreadis et al. used US county-level job-posting data from 2014 to 2023 and examined predictors of AI job growth.
In Italian cities, low-skill employment fell while high-skill employment expanded and overall labor-force participation remained steady.
The paper summarizes Capello and Lenzi's urban findings from Italian NUTS-3 data for 2009–2019.
The labor-market value of multilingual competencies varies by occupational sector: multilingualism is a clearer professional asset in client-facing, international, and language-services occupations, but its recognition is more inconsistent in technical and administrative fields with entrenched standardized, monolingual communication norms.
Synthesis of sector-specific literature on multilingualism and labor-market recognition, citing Roberts (2013) and Wilczewski and Alon (2023).
The reemployment gains among pessimistic job seekers were accompanied by a higher incidence of fixed-term contracts.
Randomized experiment linking intervention assignment to subsequent employment contract outcomes.
AI system integration is expected to increase demand for design, systems engineering, and governance skills, including AI Architects, while reducing demand for some routine tasks.
Conceptual labor-market projection based on expert synthesis; no employment or wage data were presented.
AI exposure was significantly associated with informal employment in the simulated panel.
Chi-square test of independence on the simulated microdata; χ²(2) = 234.90, p < .001, with Cramér's V = 0.217.
AI progress alone implies neither inevitable mass unemployment nor guaranteed abundance; employment and distributional outcomes are institutional equilibria shaped by governance, liability, incentives, and bargaining.
Theoretical and institutional synthesis rather than primary empirical estimation; outcomes are modeled as depending on adoption, workflow design, demand, apprenticeship, and rent allocation.
Workers ages 22–25 experienced a 6% decline in employment from late 2022 to July 2025 in the highest AI-exposure quintiles, while employment for older workers continued to grow.
ADP payroll-record study summarized in the paper; the comparison concerns employment trends across age groups and AI-exposure quintiles. The underlying study sample size is not stated.
Climate change is frequently overstated as a primary driver of migration; environmental stressors generally operate alongside socioeconomic, institutional, and developmental factors.
Interdisciplinary, sustainability-oriented synthesis drawing on comparative and case-focused analysis, literature review, and policy/governance analysis.
AI-driven hiring has mixed effects on employment outcomes: applicants aligned with algorithmic signals may benefit from faster matching, while many fresh graduates and first-time job seekers may experience lower callback and hiring rates without targeted support.
The paper presents heterogeneous employment effects as a central implication, but does not report callback or hiring-rate estimates or a study sample.
If interoperable and trusted, micro-credentials can reduce labour-market search frictions and improve matching by making discrete, job-relevant skills more visible; without standards, they may increase noise and worsen mismatches.
Economic interpretation and implications drawn from the review's findings on employer signalling, interoperability, and trust; the claim is conditional and no causal estimate is reported.
The reported retention benefits of Purdue's Course Signals may have been driven by selection effects rather than by the intervention itself.
The paper cites methodological criticism that students taking more Course Signals courses were, by construction, students who persisted longer, creating a selection effect.
AI-enabled platform economies may displace some paid tasks while creating new forms of platform-mediated work and increasing pressures on downstream creator monetisation.
Conceptual assessment of labour dynamics in AI-enabled platforms; the paper does not report a causal estimate or quantified employment effect.
The employment effects of AI are heterogeneous and depend on institutions, tasks, adoption, and time horizon rather than implying either harmlessness or certain mass displacement.
The paper compares studies reporting employment growth in higher-skilled European occupations, negative employment effects across US commuting zones, differing exposure and complementarity across demographic and country groups, and small early effects on earnings and hours.
The evidence contains substantial heterogeneity across studies, with some studies reporting negative employment effects from automation and robot adoption and others reporting positive outcomes related to productivity, wages, or skill upgrading.
Synthesis of estimates from 19 empirical studies using a three-level random-effects meta-analytic model that separates within-study and between-study variation.
The simulated ~60% gains are on par with improvements of 22–75% reported in the literature.
Comparison between the simulation results and the improvements reported in the surveyed Management Science papers (literature review lists reported improvements ranging 22–75%).
Rising aggregate output and falling labour demand can occur simultaneously — productivity gains from AI agents do not necessarily lead to stable employment.
Model results showing scenarios in which output increases while labour demand falls (analytical examples and comparative statics); theoretical analysis presented in the paper; no empirical data.
The gendered effects of AI exhibit pronounced institutional variations across different developmental stages and gender-structure conditions.
Heterogeneity/subgroup analyses reported in the paper using the 58-country panel (2000–2022) across different development stages and gender-structure conditions.
AI exhibits a significant interaction with the labor force gender structure: in scenarios of severe gender imbalance, AI's skill-restructuring effect partially mitigates the adverse impacts on female employment and economic contributions.
Empirical interaction analysis in the 58-country panel (2000–2022) and theoretical model suggesting heterogeneous effects depending on gender structure.
The United States labor market exhibits a persistent coexistence of high job vacancy rates and prolonged unemployment duration, a pattern that standard labor market theory struggles to explain.
Statement/summary of observed macro labor-market patterns cited in the paper (literature background; no new empirical sample reported in the excerpt).
The transition to the AI-integrated equilibrium is non-monotonic: the economy experiences a temporary ecological collapse driven by search frictions and delayed skill adaptation, followed by selective recovery.
Dynamical analysis using a mean-field evolutionary system and a calibrated agent-based model with bounded rationality (simulation results and theoretical dynamics).
AI's impact on employment reveals complex characteristics of both movement and design.
Statement in paper summarizing findings from the review and event-study analyses (phrase 'movement and design' appears in the text; no clarifying operationalization or empirical metrics provided in the excerpt).
The rise in productivity has had a significant impact on the labor market.
Claim based on literature reviews and event-study methods referenced by the paper; no specific datasets or sample sizes are reported in the abstract/summary provided.
AI has important implications for the future of work, creativity, and organizational learning.
Conceptual discussion and illustrative cases within the chapter regarding AI's effects on work, creative processes, and learning within organizations.
Digital transformation—encompassing automation, AI, and diffusion of digital technologies—has affected the Hungarian labour market.
Narrative synthesis of Hungarian and international academic literature, policy reports, and statistical analyses cited in the chapter (no single primary dataset or sample size reported in the summary).
Patterns suggest agencies with greater AI-susceptible occupations experience reallocation rather than displacement.
Empirical patterns in administrative employment data (2019–2024) indicating composition changes (role reallocation) in agencies with higher AI exposure, interpreted as reallocation rather than net job displacement.
The paper documents systematic associations between agencies’ concentrations of AI-exposed occupations and employment dynamics in U.S. federal agencies from 2019–2024 using administrative employment data.
Analysis of administrative employment records for U.S. federal agencies covering 2019–2024, correlating fixed occupational AI-exposure scores with observed employment dynamics at the agency level.
Current academic discussions frequently fluctuate between technological optimism focused on productivity enhancements and technological pessimism highlighting extensive job displacement.
Literature-review style claim in the paper describing the prevailing academic discourse; no specific empirical study or meta-analysis reported.
The rapid growth of artificial intelligence (AI) as a general-purpose technology is fundamentally changing the meaning of work, labour markets, and organizational structures.
Statement in paper's framing/introduction asserting broad, conceptual impact of AI; theoretical argumentation rather than empirical measurement.
The associations between automation technologies (robots and AI) and labor-market/production outcomes vary across sectors and institutional contexts in Western European and Central and Eastern European economies.
Heterogeneity analyses across sectors and regional/institutional groupings within the 32-country, 18-industry panel (Western Europe vs Central and Eastern Europe) showing differing associations by sector and regional context.
These findings contribute post-2022 information from the Swedish labor market and suggest generative AI may be associated with short-term labor market disruptions, even if long-run productivity gains and task reallocation may offset these impacts.
Empirical DiD result from monthly Swedish occupational-sector unemployment data (22 sectors) around ChatGPT release; interpretation offered by author acknowledging potential longer-run offsets.