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
Institutional determinants (data governance, standards, public infrastructure) materially influence AI diffusion and should be incorporated explicitly into diffusion models alongside human capital and capital-cost channels.
Cross-country trend comparisons and institutional analysis demonstrating correlations between institutional variables and adoption/diffusion patterns; theoretical synthesis.
Policy recommendations include investing in workforce reskilling, promoting interoperability and data portability, designing proportional risk-based regulation, using regulatory sandboxes and staged deployment, and supporting capacity building for low- and middle-income countries to avoid an AI divide.
Synthesis of policy analysis, sectoral findings and normative recommendations derived from the comparative review and gap analysis.
AI adoption can raise firm- and sector-level productivity, potentially lifting aggregate output; measuring AI’s contribution requires new indicators of 'AI intensity'.
Economic reasoning and review of literature; recommendation for measurement approaches (software/hardware investment, AI talent, use of AI services). No primary empirical measurement provided.
Regulatory design should be context-sensitive and ethics-grounded rather than one-size-fits-all.
Normative evaluation and synthesis of governance frameworks and identified gaps across jurisdictions; policy recommendations grounded in ethical principles (transparency, fairness, accountability, human rights).
AI capabilities (learning, reasoning, perception, NLP) are being integrated rapidly across healthcare, finance, education, transportation, security and justice, producing major efficiency and service-quality gains.
Sectoral case studies and documented examples cited in policy/regulatory texts and secondary literature; comparative analysis of deployments across the listed sectors.
AI is driving large productivity and capability gains across sectors.
Synthesis of sectoral case studies and secondary literature across healthcare, finance, education, transportation, security and justice; comparative policy and regulatory analysis of documented AI deployments. No large-scale primary quantitative impact evaluation reported.
International certification protocols tied to explainability and safety standards would influence investment incentives and market structure.
Policy and economic analyses in the literature synthesis arguing how standards/certification shape firm behavior and investment; no empirical causal estimation provided.
A tiered risk-management framework that allocates governance intensity to interventions by clinical criticality and autonomy is recommended to maximize benefits while containing harms.
Authors' policy recommendation derived from literature synthesis of governance frameworks, risk analyses, and implementation studies; prescriptive rather than empirically validated in large-scale trials.
Federated learning and privacy-preserving collaboration can combine data advantages without centralizing sensitive records and may reduce duplicated validation costs over time.
Technical literature and pilot studies on federated learning and privacy-preserving methods summarized in the paper; limited large-scale, long-term deployment evidence noted.
Centralized updates and monitoring by vendors can reduce operational burden for healthcare providers.
Comparative analyses and deployment reports contrasting vendor-managed services with self-managed open-source deployments; synthesized evidence and stakeholder commentary.
Open-source models enable customization and local retraining that can align models with institutional workflows and patient populations.
Cross-disciplinary literature synthesis and case reports describing local retraining/customization practices; comparative analyses of model adaptability. Evidence is drawn from diverse deployments rather than controlled trials.
Clear, harmonized regulation and procurement strategies can stimulate domestic AI suppliers, reduce dependency on foreign vendors, and capture more local economic value.
Policy analysis and market-structure discussion in the review, supported by international comparisons (e.g., Singapore, EU) and procurement case studies cited among supplementary documents.
Prioritizing AI for primary care and diagnostic applications can yield high-value health returns (reduced morbidity, earlier treatment) and improve system efficiency.
Synthesis of clinical application studies and health-economics literature within the 2020–2025 review timeframe; specific quantified returns were not uniformly reported across primary sources in the summary.
Public investment in digital health infrastructure (broadband, cloud/edge compute, interoperable data systems) is a precondition for scalable returns from AI; underinvestment will dampen both health and economic gains.
Economic and systems analysis presented in the review, drawing on international benchmarking and health-economics literature; arguments are analytical and based on modeled or literature-supported relationships rather than specified local experimental data.
AI for diabetic retinopathy screening reported an accuracy of approximately 89.3% in reviewed studies.
Reported summary statistic drawn from diagnostic performance studies identified in the 2020–2025 literature review; exact primary study sample sizes and study designs not provided in the summary.
Indonesia has demonstrated strong clinical efficacy of AI in healthcare, notably in diagnostics, telemedicine, and chronic disease management.
Narrative synthesis of literature (2020–2025) and thematic analysis of studies and pilot programs included in the review; sources include PubMed, Google Scholar, Garuda, SINTA, and 42 supplementary documents (national policy papers, SATUSEHAT governance reports, Delphi consensus studies). Specific primary study details (sample sizes, study designs) vary by application and are not uniformly reported in the synthesis.
There is a need for standards on provenance, licensing, and security auditing of AI-generated code, and potential roles for certification and liability frameworks.
Policy recommendation grounded in the identified IP, licensing, and security gaps from the literature synthesis.
Firms have strong incentives to integrate LLMs into development pipelines and to invest in internal guardrails and retraining.
Observed adoption patterns, case studies, and economic inference from potential productivity gains and risk mitigation needs presented in the review.
Human oversight and continued emphasis on computational thinking should be preserved alongside AI tool use.
Pedagogical literature and synthesis of limitations showing AI can produce plausible-but-wrong outputs and that human reasoning mitigates risks.
Rigorous verification, QA protocols, and security audits are necessary when integrating AI-generated code into production systems.
Cross-study synthesis and case analyses indicating nontrivial defect and vulnerability rates in AI outputs and the costs/remediation steps observed in practice.
Generative AI tools lower entry barriers for novices and can speed learning of programming tasks.
Pedagogical assessments and user studies comparing novice performance and learning speed with and without AI assistance, as reported in the literature synthesized by the paper.
The most promising deployment mode is augmentation (AI suggestions plus human oversight) rather than full automation.
Cross-study synthesis of user studies and case studies showing improved outcomes when humans review and modify AI outputs and failures when relying on fully automated outputs.
Large language models (LLMs) can accelerate coding tasks, debugging, and documentation, functioning effectively as collaborative coding assistants.
Synthesis of multiple user studies and productivity measurements (task completion time, workflow observations) and code-generation benchmarks reported in the reviewed empirical literature.
Policy instruments that merit evaluation include retraining programs, wage insurance, R&D subsidies, tax incentives for productive AI adoption, and competition policy for AI platforms to smooth transitions and share gains.
Policy recommendations synthesized from reviewed literature and institutional reports; the paper calls for evaluation but does not provide new experimental or quasi‑experimental evidence on these instruments.
Realizing net social gains from AI/robotics requires strategic public policy, ethical regulation, investment in skills and data infrastructure, and inclusive innovation strategies.
Policy prescription based on synthesis of cross‑study findings and normative analysis; recommendations draw on secondary evidence about risks and opportunities but are not themselves empirically validated within the paper.
In India, AI/robotics are transforming manufacturing, healthcare, agriculture, infrastructure, and smart cities, enabling data‑driven policy and business decisions and offering potential for sustainable development and inward investment.
Country case studies and sectoral examples from secondary reports focused on India (multilateral and consulting firm studies); descriptive evidence rather than causal estimation; sample sizes and empirical details vary by source and are not summarized quantitatively in the paper.
Adoption of AI/robotics influences major macroeconomic indicators (GDP growth, capital flows, productivity metrics) and attracts foreign investment.
Descriptive analysis using secondary macro indicators and cited studies/reports from multilateral organizations and consulting firms; evidence is correlational and heterogeneous across studies; specific sample sizes vary by cited source and are not consolidated in the paper.
AI and robotics automate routine and labour‑intensive tasks, lower unit costs, reduce errors, and raise output quality and throughput across manufacturing, services, healthcare, agriculture, and infrastructure.
Sectoral adoption examples and sector reports summarized in a qualitative literature review (secondary sources from industry reports and multilateral organizations); no pooled quantitative meta‑analysis or uniform sample size reported.
AI and robotics are driving a renewed productivity and growth phase across industries, raising GDP, capital productivity, and competitiveness.
Qualitative literature synthesis and descriptive analysis of secondary macro indicators and sectoral examples drawn from reports by international institutions and consulting firms; no original causal estimation; sample sizes and effect magnitudes not reported in the paper.
Adoption of generative neural-network–based audiovisual AI is likely inevitable and will significantly raise productivity in content creation.
Narrative review and conceptual synthesis of secondary literature on generative neural networks and industrial/market analyses; no new primary data collected (methodology section explicitly states secondary-data narrative review).
Firms are likely to invest in proprietary datasets, model-locking, certification/verification services, insurance, and compliance/legal risk management, which will influence adoption timing and scale.
Strategic behavior analysis in the review supported by referenced industry behavior and economic incentives; no firm-level empirical investment data or sample sizes provided.
Generative audiovisual models promise large productivity gains in content creation (lower marginal costs and faster content production).
Economic reasoning and secondary literature cited in the review; no primary quantitative measurement or sample size reported in the paper.
Practical research directions include: studying platformization impacts on informal labor and small suppliers using causal designs; combining satellite imagery with ML to measure resource flows and supply-chain disruptions linked to market outcomes; developing ML methods robust to intermittent data and structural breaks; and evaluating AI-enabled policies (credit scoring, logistics routing, demand forecasting) through pilots and RCTs to measure welfare and distributional effects.
Paper's concluding/practical recommendations synthesised from literature; no empirical pilots/results presented in the paper.
Cost-effective, explainable AI models are preferred in African OSCM contexts where computational resources and technical capacity are limited.
Design recommendation from the paper's discussion on resource constraints and capacity.
AI policies and algorithmic accountability mechanisms must be tailored to weak institutional environments, for example by leveraging community norms when formal legal enforcement is limited.
Normative recommendation in the paper based on institutional analysis and literature review.
Algorithmic and policy design in African OSCM contexts should account for informal-contract enforcement, cash-based transactions, and heterogeneous preferences rather than assuming strong formal enforcement and homogeneous agents.
Policy and design implications drawn conceptually from the paper's synthesis of institutional and market features.
Recommended empirical methods for African OSCM and AI economics research include combining causal inference designs (RCTs, natural experiments, IV) with structural modeling, simulation, transfer learning, domain adaptation, and robustness checks to handle small or nonrepresentative datasets.
Methodological guidance in the paper derived from cross-disciplinary literature.
Useful data sources for AI economics research in African OSCM contexts include mobile-phone metadata, fintech/platform transaction logs, household/business surveys, administrative records, satellite/remote sensing, and crowdsourced field data.
Practical data recommendations from the paper's methodological discussion.
Abundant natural resources but low economic outcomes motivate AI-assisted monitoring (satellite imagery), predictive models for value-chain improvements, and incentive/contract design to address extraction externalities.
Conceptual proposal tying resource economics and AI applications in the paper.
High environmental constraints (limited infrastructure, frequent shocks) motivate the development and testing of robust, low-data, low-compute AI methods for supply-chain optimization, demand forecasting, and inventory management.
Paper's synthesis linking environmental constraints to methodological needs for AI in OSCM.
Weak formal institutions alongside strong informal norms allow researchers to investigate how algorithmic interventions (automated enforcement, marketplaces, credit scoring) interact with informal governance and trust networks.
Conceptual mapping from institutional theory to algorithmic governance literature in the paper.
Africa’s large informal sectors function as a laboratory to study how AI-driven automation, platform markets, and pricing algorithms affect informal firms and workers (displacement, complementarities, informal-contract dynamics).
Conceptual linkage between informal-economy characteristics and AI/economics research opportunities described in the paper.
The authors recommend leveraging diverse data sources (administrative records, surveys, behavioral data, remote sensing) and mixed-methods designs for future empirical work on African OSCM contexts.
Methodological recommendations in the paper based on literature synthesis.
Managing institutions (interplay of formal and informal governance, regulation, trust mechanisms) in Africa provides fertile ground for advancing institutional theories in OSCM.
Institutional economics and governance literature synthesized in the paper.
Managing environmental hostility (resilience, adaptation to shocks, infrastructure limitations) in African contexts can drive OSCM theory on resilience and adaptation strategies.
Literature review on shocks, resilience, and infrastructure constraints; conceptual proposal.
Managing resources in African supply chains (resource extraction, allocation, quality gaps) highlights unique allocation problems and quality-related frictions for OSCM theory.
Conceptual argument drawing on resource economics and supply-chain literature.
Serving consumer markets in Africa (distribution, last-mile delivery, demand heterogeneity) offers opportunities to study distinct distribution models and last-mile challenges.
Conceptual mapping from literature on market structures and logistics in African contexts.
Five OSCM research themes where African contexts can advance theory are: serving consumer markets, managing resources, managing factor market rivalry, managing environmental hostility, and managing institutions.
Framework developed through literature synthesis in the paper; no empirical validation provided.
Levers such as reducing training costs, improving perceived safety, and targeted marketing can shift the system toward a positive adoption equilibrium.
Simulation-based sensitivity analysis reported in Essay 2 that identifies how changes in parameters alter basins of attraction and increase likelihood of the favorable equilibrium (no field experiment or empirical intervention evidence provided).
Simulations show behavior can converge to an 'ideal equilibrium' in which owners, employees, and customers all accept service robots.
MATLAB simulations of the three-player evolutionary game that trace dynamic behavior under specific parameterizations and initial conditions (details of parameter values and number of simulation runs not provided in summary).