Evidence (6917 claims)
Adoption
8625 claims
Productivity
7686 claims
Governance
6917 claims
Human-AI Collaboration
6574 claims
Org Design
4189 claims
Innovation
4131 claims
Labor Markets
3588 claims
Skills & Training
2985 claims
Inequality
2066 claims
Evidence Matrix
Claim counts by outcome category and direction of finding.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 761 | 200 | 101 | 904 | 2020 |
| Governance & Regulation | 829 | 400 | 191 | 122 | 1566 |
| Organizational Efficiency | 784 | 193 | 125 | 84 | 1197 |
| Technology Adoption Rate | 637 | 236 | 124 | 97 | 1103 |
| Research Productivity | 431 | 131 | 58 | 340 | 972 |
| Output Quality | 481 | 183 | 59 | 47 | 770 |
| Decision Quality | 332 | 177 | 82 | 49 | 647 |
| Firm Productivity | 439 | 57 | 88 | 20 | 610 |
| AI Safety & Ethics | 218 | 279 | 66 | 33 | 602 |
| Market Structure | 181 | 170 | 123 | 24 | 503 |
| Task Allocation | 214 | 64 | 72 | 33 | 388 |
| Skill Acquisition | 174 | 62 | 62 | 17 | 315 |
| Innovation Output | 204 | 27 | 45 | 18 | 295 |
| Employment Level | 105 | 54 | 108 | 13 | 282 |
| Fiscal & Macroeconomic | 132 | 69 | 43 | 26 | 277 |
| Consumer Welfare | 117 | 63 | 42 | 11 | 233 |
| Firm Revenue | 154 | 48 | 26 | 3 | 231 |
| Task Completion Time | 173 | 31 | 8 | 12 | 225 |
| Inequality Measures | 44 | 123 | 50 | 6 | 223 |
| Worker Satisfaction | 89 | 65 | 22 | 12 | 188 |
| Error Rate | 71 | 92 | 10 | 2 | 175 |
| Regulatory Compliance | 77 | 69 | 14 | 5 | 165 |
| Automation Exposure | 58 | 56 | 26 | 13 | 156 |
| Training Effectiveness | 96 | 21 | 14 | 19 | 152 |
| Wages & Compensation | 77 | 37 | 25 | 6 | 145 |
| Team Performance | 86 | 17 | 27 | 10 | 141 |
| Developer Productivity | 95 | 17 | 14 | 6 | 133 |
| Job Displacement | 12 | 81 | 21 | 1 | 115 |
| Hiring & Recruitment | 52 | 7 | 8 | 3 | 70 |
| Creative Output | 32 | 20 | 8 | 3 | 64 |
| Skill Obsolescence | 5 | 47 | 6 | 1 | 59 |
| Social Protection | 28 | 16 | 8 | 2 | 54 |
| Labor Share of Income | 17 | 19 | 17 | — | 53 |
| Worker Turnover | 11 | 12 | — | 3 | 26 |
| Industry | — | — | — | 1 | 1 |
Governance
Remove filter
Priority investments should target computational infrastructure, local model validation capacity, and training for clinicians and data scientists to increase adoption and trust in synthetic-data–supported AI.
Implementation and capacity-building analyses from the reviewed literature highlighting gaps in infrastructure, validation capability, and human capital; recommendation-based evidence rather than new empirical trials.
Vendor support, warranties, and service-level agreements (SLAs) are important for clinical adoption and liability management.
Policy and implementation literature, industry reports, and stakeholder feedback synthesized in the paper highlighting the role of vendor contractual commitments in adoption decisions.
Proprietary systems lead on reliability, maintenance, and validated integrations with clinical systems.
Literature synthesis including vendor case studies, deployment reports, and stakeholder surveys indicating more mature productization and validated integrations for proprietary offerings.
Open-source deployment options (e.g., on-premises) reduce data-sharing exposure and improve privacy.
Aggregated evidence from deployment reports and technical papers describing on-premises and local inference architectures; industry analyses of data governance tradeoffs.
Open-source models provide greater transparency and inspectability, enabling better auditability and explainability.
Systematic literature synthesis of peer-reviewed studies, industry reports, and case studies comparing open-source and proprietary systems; comparative analysis highlights inspectability of open-source code/models. No new primary experiments reported.
Coordinated policy reform, targeted infrastructure investment, workforce training, and equity-focused implementation are strategic priorities to realize AI’s potential in Indonesian healthcare.
Consensus recommendations drawn from the narrative synthesis, thematic analysis, and Delphi consensus studies included among the 42 supplementary documents and the broader 2020–2025 literature body.
Recommended research priorities for economists include measuring how adoption changes task mixes and wages, quantifying verification/remediation costs, estimating productivity gains net of security/IP costs, and studying market dynamics from centralized model providers.
Author recommendations based on identified gaps in the empirical literature synthesized by the paper.
Recommended policy levers include data-governance rules, provenance and watermarking standards, liability frameworks, copyright clarifications, competition policy, and taxes/subsidies to internalize externalities.
Policy recommendations synthesized from legal, regulatory, and economic literatures within the review; presented as qualitative guidance rather than tested policy interventions.
A structured three-stage framework (input/process/output) clarifies where different risks and regulatory rules apply to generative audiovisual systems.
Framework presented in the paper as a conceptual synthesis of reviewed literatures; supported by cross-references to legal, technical, and ethical sources within the review.
The paper introduces IJOPM’s Africa Initiative (AfIn) to support Africa-based OSCM research, outlining motivation, objectives, review process, and researcher support mechanisms.
Descriptive account within the paper (administrative/initiative description rather than empirical evidence).
High‑frequency sensor and satellite data, processed with AI, improve precision in measuring yields, input use, and environmental externalities, enhancing the quality of economic impact evaluations and policy targeting.
Methodological and validation studies using high‑resolution satellite imagery and field sensors that show improved measurement accuracy versus traditional survey methods; referenced empirical demonstrations in the literature.
The paper proposes specific metrics and empirical follow-ups (e.g., generation-to-verification throughput ratios, defect accumulation rates, time-to-acceptance for machine-generated artifacts, incident rates attributable to unverified AI outputs) to validate the model.
Explicit recommendations and measurement proposals listed in the paper; no empirical implementation provided.
Recommended next steps include building and calibrating ABMs with agent heterogeneity, prototyping technical implementations of token verification (proof-of-query receipts, cryptographic attestation), and red-teaming for spoofing/evasion.
Paper's research & policy next-steps and operational recommendations; no implementation results included.
Chain-of-Thought prompts/internal reasoning simulate richer, multi-step decision processes in agents compared with conventional single-step decision rules.
Methodological description: use of CoT prompts/internal reasoning to model multi-step deliberation in agents. This is a documented implementation detail and conceptual claim in the paper.
The framework replaces static, rule-based agent decision-making with LLM-powered cognitive agents that perceive environment signals, deliberate using Chain-of-Thought, and act—without hand-coded behavior rules.
Model architecture description: each agent is an LLM-driven cognitive unit implementing the PDA loop; explicit statement that behavior is not hand-coded but emerges from language-model deliberation. This is a design/implementation claim rather than an empirical result.
Team Situation Awareness (shared perception, comprehension, projection) remains a useful analytic anchor for HAT even with agentic AI.
Conceptual analysis mapping Team SA components onto agentic AI interactions; literature review of Team SA utility in HAT contexts.
DAR produces ten falsifiable propositions explicitly mapped to measurement constructs, making the framework empirically testable.
Derivation and listing of ten testable propositions in the paper, each linked to observable measures and prioritized by feasibility. Theoretical derivation, no empirical tests provided.
Common uses of AI among practitioners include generating code snippets, suggesting fixes, accelerating routine tasks, surfacing design patterns or documentation, and scaffolding prototypes.
Practice-focused qualitative data from interviews and workflow analysis at Netlight; authors list these use-cases as commonly reported by practitioners; frequency counts not provided.
Practitioners use AI primarily as a practical assistant (coding, debugging, prototyping, knowledge retrieval) rather than as a fully autonomous developer.
Reported practitioner accounts and observations from the Netlight field study (interviews/observations); examples of tasks AI is used for were documented in the paper; sample limited to experienced consultants at one firm.
Experienced IT professionals at Netlight are already integrating AI tools into everyday development work.
Qualitative field study conducted at Netlight Consulting GmbH using interviews, observations, and analysis of practitioner workflows; single-firm sample (Netlight); exact number of participants not reported.
BERT-family encoders provide superior contextual understanding for sentiment analysis, intent detection, behavioural segmentation, and feature extraction from user signals compared to simpler feature pipelines.
Use of BERT encoders for classification tasks with offline metrics reported such as classification accuracy for intent/sentiment and user embedding quality for segmentation. (Specific datasets and sample sizes are not provided.)
Automated equivalency systems require algorithmic oversight features (audit trails, human-in-the-loop checks) to maintain trust and labor-market legitimacy.
Governance recommendation following best practices in algorithmic accountability; not supported by empirical testing of oversight mechanisms in this context.
AI tools (automated document parsing/NLP, translation, equivalency-prediction classifiers, anomaly detection) can scale credential processing and reduce transaction costs and processing time.
Paper cites potential AI capabilities and application areas; the claim is inferential from known AI functionalities, with no implementation benchmark or throughput numbers provided.
Continuous monitoring and observability for performance, compliance, and drift are essential to maintain operational stability and detect model or process degradation.
Prescriptive claim grounded in engineering practice and comparative analysis of failure modes; supported by illustrative deployments; no quantitative evaluation of monitoring impact reported.
Core governance components should include policy enforcement integrated into development and deployment pipelines, risk controls for data/model behavior/automated actions, explicit human-in-the-loop and human-on-the-loop oversight, continuous monitoring/logging/incident-response, and role-based governance structures linking legal, compliance, IT, and business units.
Prescriptive design based on literature synthesis and practitioner experience; described as core components in the proposed reference pattern (conceptual, case-illustrated).
Research needs include empirically measuring prevalence and average loss from prompt fraud incidents, evaluating effectiveness and cost-effectiveness of technical mitigations (watermarking, provenance), and modeling firm-level investment decisions under varying regulatory/insurance regimes.
Authors' recommended agenda for further research based on identified gaps in the paper's qualitative analysis.
The United States manages the openness–security trade-off via a decentralized, rights‑based coordination emphasizing procedural transparency and public accountability.
Qualitative content analysis of national‑level policy texts: 18 U.S. policy documents coded across the same four analytical dimensions.
If companies are treated as recipients, they would be required to comply with nondiscrimination obligations (e.g., Title VI, Title IX, Section 504) in education contexts and may be subject to enforcement actions, corrective requirements, and private suits where applicable.
Interpretation of recipient obligations under existing civil‑rights statutes and enforcement mechanisms; doctrinal analysis and illustrative case law.
Systems biology, constraint‑based metabolic modeling (e.g., FBA), kinetic modeling, and hybrid models are effective tools to predict fluxes and identify metabolic bottlenecks.
Discussion and aggregation of modeling studies using COBRA/OptFlux frameworks, FBA simulations, and kinetic/dynamic modeling applied to engineered strains to predict flux changes and suggest genetic interventions; validated in multiple reported DBTL cycles.
Engineered microorganisms are maturing into modular, programmable “microbial factories” capable of producing complex chemicals, specialty compounds, and next‑generation biofuels.
Synthesis of multiple experimental case studies reported in the literature (bench and pilot scale fermentations) demonstrating microbial production of natural products, specialty chemicals, and biofuel molecules using engineered strains and heterologous pathways; methods include pathway assembly, enzyme engineering, and fermentation optimization.
The authors introduce clinical-model instruments such as the Model Temperament Index (behavioral profiling), Model Semiology (structured symptom lexicon), and M-CARE (standardized case reporting).
Proposed indices and reporting formats presented in the methods and applied in demonstrations/cases within the paper.
The paper proposes a five-layer diagnostic framework: staged assessment from symptom description to mechanistic localization and prognosis.
Framework design documented in the paper and applied in case demonstrations (descriptive pipeline combining symptom elicitation, profiling, semiology, imaging/localization, and reporting).
Neural MRI (Model Resonance Imaging) maps five medical neuroimaging modalities to corresponding AI interpretability techniques (e.g., structural → weight-space maps, functional → activation dynamics, connectivity → representational similarity).
Methodological mapping and toolkit design described in the paper (conceptual mapping and implemented open-source toolkit).
The authors present a discipline taxonomy comprising 15 subdisciplines grouped into four divisions: Basic Model Sciences, Clinical Model Sciences, Model Public Health, and Model Architectural Medicine.
Taxonomic synthesis produced by the authors from interpretability, reliability, governance, and architecture literatures (documented taxonomy in the paper).
The paper defines 'Model Medicine' as a unified research program treating AI models like organisms with diagnosable, classifiable, and treatable states.
Conceptual framing and theoretical synthesis presented in the paper (literature-driven argumentation; no empirical sample required).
A research agenda prioritizing empirical evaluation, model transparency, and rigorous impact assessment is required to translate conceptual promise into measurable public value.
Explicit recommendation in the blurb identifying research priorities; not an empirical claim but a proposed course of action.
Illustrative vignettes show AI in action: logistics optimization for trade, AI models for national fiscal decision-making, and algorithmic job-acceleration for individual labor market navigation.
Reference to specific case vignettes contained in the book; these are illustrative scenarios rather than empirical case studies with measured outcomes.
Ten defining policy questions structure the book’s approach, turning abstract AI capabilities into operational policy choices.
Descriptive claim about the book's organization; verifiable by inspecting the book's table of contents (no external empirical data).
The compendium issues specific policy-design recommendations for economic policymakers: deploy proportional compliance obligations and regulatory sandboxes, subsidize or certify third‑party auditors, monitor credit availability and pricing post‑implementation, and coordinate cross‑border standards.
Explicit policy recommendations listed in the "Policy design recommendations" subsection; derived from the paper's interdisciplinary analysis.
The protocol has been prepared/indexed across 15 strategic languages to facilitate international diffusion and comparative uptake.
Stated multilingual/global indexing claim in the compendium (15 languages).
The paper implements a "White Box" regulatory protocol for AI in Mexico's financial sector requiring algorithmic transparency, auditability, explainability, and non‑discrimination standards for credit/FinTech algorithms.
Output of the technical protocol described in the compendium; developed from a forensic audit of source materials and legal-methodological synthesis (doctrinal/comparative analysis).
The compendium proposes recognizing "Digital Sovereignty" as a new fundamental human right that protects individuals’ autonomy, data sovereignty, due process, and non-discrimination in algorithmic financial decision‑making.
Normative definitional claim in the protocol; grounded in the author's doctrinal and comparative legal analysis across 12 years (2014–2026).
Recommended policy approach: run pilots to empirically measure trade‑offs, combine obligations with capacity building (technical assistance, shared datasets, sandboxes), harmonize with international frameworks, and use staged implementation with cost‑benefit analyses.
Policy recommendations derived from the compendium’s interdisciplinary synthesis and economic/policy analysis (prescriptive, not empirically validated within the paper).
Policy operationalization should include algorithmic impact assessments, audit logs, disclosure regimes to regulators/judiciary, redress/grievance mechanisms, and governance principles (open, transparent, accountable).
Prescriptive policy instruments and standards proposed in the compendium based on the forensic audit and normative design work; descriptive claim about the protocol’s recommended instruments.
There is a need for standardized metrics to quantify benefits and costs of governed hyperautomation (e.g., ROI adjusted for compliance risk, incident rate per automation scale, oversight hours per automated transaction, model drift frequency and remediation cost).
Paper's recommendations and research agenda calling for standardized metrics and empirical studies; prescriptive statement rather than empirical finding.
Researchers and policymakers should promote auditable, privacy-preserving attribution standards and independent audits while supporting randomized trials and field experiments under privacy constraints.
Policy/actionable takeaways informed by methodological challenges and literature on randomized trials and privacy-preserving methods; prescriptive guidance rather than an empirically tested program.
There is a need for standardized benchmarks and privacy-preserving shared datasets to enable independent economic evaluation of ad-tech.
Methodological recommendation informed by stated data access asymmetries and reproducibility concerns; not accompanied by a new benchmark in the paper.
Antitrust analysis of ad-tech should incorporate algorithmic effects such as endogenous use of ML to entrench platform position and data network effects.
Theoretical and policy argument drawing on platform economics and ML scale advantages; recommendation rather than empirical finding.
Combining secure aggregation and differential privacy can materially reduce centralized custody risks.
Conceptual systems design and analytical discussion combining cryptographic and statistical privacy mechanisms; threat model argues joint effect reduces reconstruction and limits leakage. No field measurements of residual risk provided.
Secure aggregation protocols (cryptographic aggregation, MPC) can prevent reconstruction of individual updates and thus materially reduce risk of exposing raw behavioral logs to centralized custodians.
Systems design and threat modeling mapping secure aggregation techniques to privacy risk reduction; references to standard cryptographic protocols. Empirical support limited to conceptual mapping and prototype/simulation; no deployment measurements.