Evidence (7198 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
8921 claims
Filter claims →
Productivity
8002 claims
Filter claims →
Governance
7198 claims
Filtered →
Human-AI Collaboration
6864 claims
Filter claims →
Org Design
4398 claims
Filter claims →
Innovation
4286 claims
Filter claims →
Labor Markets
3629 claims
Filter claims →
Skills & Training
3001 claims
Filter claims →
Inequality
2141 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 | 790 | 208 | 103 | 950 | 2117 |
| Governance & Regulation | 869 | 411 | 195 | 126 | 1630 |
| Organizational Efficiency | 817 | 202 | 126 | 87 | 1243 |
| Technology Adoption Rate | 675 | 258 | 128 | 106 | 1178 |
| Research Productivity | 462 | 138 | 64 | 347 | 1023 |
| Output Quality | 501 | 193 | 61 | 52 | 807 |
| Decision Quality | 346 | 180 | 84 | 51 | 668 |
| AI Safety & Ethics | 235 | 285 | 70 | 34 | 630 |
| Firm Productivity | 452 | 58 | 91 | 20 | 627 |
| Market Structure | 184 | 171 | 123 | 24 | 507 |
| Task Allocation | 221 | 65 | 76 | 34 | 401 |
| Skill Acquisition | 176 | 62 | 62 | 17 | 317 |
| Innovation Output | 207 | 28 | 48 | 18 | 303 |
| Fiscal & Macroeconomic | 135 | 72 | 44 | 26 | 284 |
| Employment Level | 105 | 56 | 108 | 13 | 284 |
| Consumer Welfare | 121 | 67 | 45 | 11 | 244 |
| Firm Revenue | 160 | 50 | 28 | 4 | 242 |
| Task Completion Time | 182 | 33 | 10 | 13 | 239 |
| Inequality Measures | 45 | 126 | 50 | 6 | 227 |
| Worker Satisfaction | 94 | 73 | 23 | 12 | 202 |
| Error Rate | 76 | 98 | 11 | 4 | 189 |
| Regulatory Compliance | 81 | 73 | 17 | 7 | 178 |
| Automation Exposure | 61 | 59 | 26 | 14 | 163 |
| Training Effectiveness | 97 | 21 | 14 | 19 | 153 |
| Wages & Compensation | 78 | 37 | 25 | 6 | 146 |
| Developer Productivity | 105 | 18 | 14 | 6 | 144 |
| Team Performance | 87 | 17 | 28 | 10 | 143 |
| Job Displacement | 12 | 83 | 21 | 1 | 117 |
| Hiring & Recruitment | 52 | 8 | 8 | 3 | 71 |
| Social Protection | 39 | 17 | 8 | 2 | 66 |
| Creative Output | 32 | 20 | 8 | 3 | 64 |
| Skill Obsolescence | 5 | 49 | 6 | 1 | 61 |
| Labor Share of Income | 17 | 19 | 17 | — | 53 |
| Worker Turnover | 15 | 14 | — | 3 | 32 |
| Industry | — | — | — | 1 | 1 |
Governance
Remove filter
The audit surface follows the same one-versus-N pattern: DPM logs two LLM calls per decision while summarization logs 83-97 on LongHorizon-Bench.
Empirical measurement on LongHorizon-Bench reported in the paper: logged LLM calls per decision are 2 for DPM vs 83-97 for summarization.
DPM is additionally 7-15x faster at binding budgets, making one LLM call at decision time instead of N.
Empirical runtime/efficiency measurement reported in the paper (range 7-15x speedup) comparing number of LLM calls and latency under tight memory budgets.
At a 20x compression ratio, DPM improves reasoning coherence by +0.53 (Cohen's h=1.13, p=0.0034) compared to summarization-based memory (paired permutation, n=10).
Paired permutation test over 10 cases at a 20x compression ratio; reported effect +0.53 with Cohen's h=1.13 and p=0.0034.
At a 20x compression ratio, DPM improves factual precision by +0.52 (Cohen's h=1.17, p=0.0014) compared to summarization-based memory (paired permutation, n=10).
Paired permutation test over 10 cases at a 20x compression ratio; reported effect +0.52 with Cohen's h=1.17 and p=0.0014.
On ten regulated decisioning cases at three memory budgets, DPM matches summarization-based memory at generous budgets and substantially outperforms it when the budget binds.
Empirical evaluation on 10 decisioning cases across three memory budgets; comparison between DPM and summarization-based memory as reported in the paper (n=10).
We propose Deterministic Projection Memory (DPM): an append-only event log plus one task-conditioned projection at decision time.
Method/architectural proposal described in the paper.
Presumptuousness in legal AI is systematic but addressable, and addressing it is a necessary step towards systems that reliably support, rather than supplant, human judgment wherever decisions must await sufficient evidence.
Synthesis conclusion in paper based on the benchmark experiments, comparisons across prompting methods, and SPEC results.
SPEC achieves 89% overall accuracy, while appropriately deferring when evidence is insufficient.
Empirical evaluation of SPEC reported in paper: overall accuracy reported as 89% and behavior of proper deferral on insufficient-evidence cases.
We introduce SPEC (Structured Prompting for Evidence Checklists), a structured framework requiring explicit identification of missing information before any determination.
Methodological contribution described in paper: new prompting/framework (SPEC) that enforces explicit missing-information identification prior to decision.
Through a collaboration with the Colorado Department of Labor and Employment, we secured access to official training materials and guidance to design a novel benchmark that systematically varies information completeness.
Methodological description in paper: collaboration with state agency and dataset/benchmark construction using official training materials and guidance.
The framework is applied to Canada's 2025-2026 national AI Strategy consultation with n = 5,253 respondents across two independent policy topics.
Empirical application reported in the paper; dataset description gives sample size and two policy topics.
This paper introduces 'participatory provenance': a measurement framework grounded in optimal transport theory, causal inference and semantic analysis that tracks how individual public submissions are transformed, filtered or lost through AI-mediated summarization.
Methodological contribution described in the paper (framework design combining optimal transport, causal inference, semantic analysis).
Local governments should develop coordinated AI policy mixes, align differentiated policy pathways with regional conditions, and prioritize technology R&D support, talent cultivation and collaboration, and application demonstration and promotion to sustain long-term regional competitiveness.
Authors' policy recommendations derived from the fsQCA findings and interpretation of which conditions are recurrent/core across configurations.
Technology R&D support, talent cultivation and collaboration, and application demonstration and promotion are the most recurrent core policy conditions across the identified configurations.
Frequency/core-condition analysis within the fsQCA configurations reported by the authors showing these three policy instruments repeatedly appear as core conditions.
The study identifies three driving pathways to sustained competitiveness: (supply and demand)-environmental resonance; demand-driven (supply-environmental) assurance; and supply–demand complementarity, which together cover five specific configurations.
Reported fsQCA solution paths (three aggregated driving pathways and five specific configurations) derived from the analysis of provincial AI policy instruments.
Sustained competitiveness is achieved through multiple equivalent configurations of policy instruments (i.e., policy instrument combinations rather than single instruments).
fsQCA results reported in the paper showing multiple configurations (solution paths) that are associated with high regional competitiveness.
AI systems currently provide more consistent fraud warnings than lay humans in an identical advisory role.
Aggregate comparison from the preregistered experiment showing humans had nonzero endorsement and higher suppression rates while all tested LLMs showed 0% endorsements and lower suppression under pressure (human n=1,201; AI conversations n=3,360).
Human advisors endorsed fraudulent investments at baseline rates of 13-14%.
Human benchmark of 1,201 participants run in the preregistered experiment; reported baseline endorsement rates for fraudulent scenarios.
Motivated investor framing did not suppress AI fraud warnings; if anything, it marginally increased them.
Preregistered experiment across seven leading LLMs and twelve investment scenarios; 3,360 AI advisory conversations analyzed comparing motivated vs. baseline investor framings.
Future research should prioritize hybrid human-AI decision frameworks, robust evaluation in diverse emerging market contexts, and development of regulatory technology solutions that balance innovation with systemic stability.
Recommendations and Conclusion section derived from identified gaps and themes in the scoping review.
AI-driven approaches show substantial promise for enhancing financial risk management in emerging markets, particularly in credit scoring, fraud detection, and market forecasting.
Overall conclusion synthesizing reported improvements and application areas across the 64 studies; qualitative and quantitative findings summarized by authors.
Neural networks and ensemble methods demonstrate superior predictive accuracy compared to traditional methods.
Synthesis of comparative results across included studies indicating better predictive performance of neural networks and ensemble methods in market prediction, credit scoring, and related tasks.
Performance improvements (of AI methods) range from 15% to 35% over traditional methods.
Aggregate statement in Results summarizing reported performance improvements across reviewed studies (no single-trial RCT; based on comparative performance metrics reported by included studies).
This work provides a replicable methodology for auditing institutional ML systems and highlights the importance of evaluating construct validity alongside statistical fairness.
Paper presents the ASP-HEI Cycle-informed replica-based audit method and argues for assessing construct validity in addition to statistical fairness metrics.
We evaluate disparities by gender, age, and residency status across the full pipeline (training data, model predictions, and post-processing) using standard fairness metrics.
Paper reports conducting evaluation across the full ML pipeline using standard fairness metrics disaggregated by gender, age, and residency status.
We present a replica-based audit of a deployed Early Warning System (EWS), replicating its model using institutional training data and design specifications.
Statement in paper describing a replica-based audit using Centennial College's institutional training data and the system's design specifications; multi-year collaboration and prior ethnographic work informing approach.
In the AI era, digital sovereignty is more plausibly pursued through institutionally governed interdependence than through technological autonomy.
Normative/conclusive argument presented by the paper (theoretical recommendation). This is an argumentative conclusion rather than an empirically demonstrated finding in the provided text.
The sovereign SLM+RAG configuration is discussed as one possible operational pathway through which the Governance Membrane architecture may be instantiated in contexts where embedded-mode governance is feasible.
Specific implementation pathway proposed/discussed by the authors (design suggestion). No empirical testing or sample information provided in the supplied text.
As a secondary, design-oriented contribution, the paper proposes the Governance Membrane as a reference architecture for operationalizing the Governed Interdependence paradigm, and introduces the Normative Compliance Model, the Infrastructure Status Index, and the Cognitive Dependence Index as complementary instruments for normative alignment and governance calibration.
Design-oriented conceptual proposal described in the paper (framework/instrument design). No empirical evaluation or sample details reported in the provided text.
The paper develops the Governed Interdependence paradigm, which reconceptualizes digital sovereignty as the institutional capacity to govern structured participation in globally distributed AI infrastructures rather than to achieve full technological autonomy.
Primary theoretical contribution described in the paper (conceptual/model development). This is a proposed framework introduced by the authors rather than an empirically validated result.
The paper provides firm-level empirical evidence from an underexplored emerging market context (Nigerian listed firms) on the relationship between AI adoption in financial reporting and audit quality.
Study sample and context are Nigerian listed firms; empirical analyses (content analysis, archival audit data, SEM) reported in the paper.
The study operationalizes AI adoption using a disclosure-based AI adoption index, representing a methodological advancement for measuring firm-level AI adoption in financial reporting.
Content analysis of corporate annual reports used to construct a disclosure-based AI adoption index; index applied in SEM analysis.
The positive relationship between AI adoption and audit quality is partially mediated by improvements in internal control quality.
SEM mediation analysis including internal control quality as a mediator; internal control quality measured through disclosure/content analysis and related archival indicators; audit quality captured via restatements and audit fees.
The positive relationship between AI adoption and audit quality is partially mediated by improvements in reporting transparency.
SEM mediation analysis including a reporting transparency measure derived from content analysis of annual reports; archival audit data used for audit quality indicators.
AI adoption is positively associated with audit quality in Nigerian listed firms.
Mixed-method quantitative design combining content analysis of corporate annual reports (to construct a disclosure-based AI adoption index) and archival audit data; Structural Equation Modeling (SEM) used to test the direct relationship. Audit quality modeled as a latent construct reflected by financial restatements and audit fees.
Sustainable development outcomes in MENA economies are driven not only by technology adoption but by the interaction between digital infrastructure, AI, and institutional readiness.
Regression models including interaction terms between digital transformation, AI measures, and indicators of institutional readiness within the System GMM analysis.
There is significant regional heterogeneity: Gulf Cooperation Council (GCC) countries exhibit stronger effects of digital transformation and AI on sustainable development than non-GCC MENA economies.
Subgroup/interaction analyses by region (GCC vs non-GCC) within the System GMM framework reported differential coefficients.
Artificial intelligence (AI) has a positive but weaker impact on sustainable development relative to digital transformation, reflecting its complementary and maturity-dependent role within the digital ecosystem.
Same System GMM regressions on panel of MENA economies (2010–2023) that include measures of AI and digital transformation; reported positive but smaller coefficient for AI.
Digital transformation is the primary driver of sustainable development in MENA economies, exerting a stronger and more consistent effect than AI.
Dynamic panel data analysis of MENA economies (2010–2023) using System GMM; reported comparative effect sizes of digital transformation vs. AI in regression results.
The authors propose corresponding analytical extensions to the framework to address the three structural breaks in agentic systems.
Paper presents proposed analytical extensions (methodological proposals) tied to each identified structural break.
Cross-architecture comparison reveals a governance coverage gradient: deterministic rule engines achieve full DES-property fillability.
Analytic cross-architecture comparison reported in the paper (comparative analysis across four architectures); deterministic rule engines identified as achieving 'full' fillability of DES-properties.
The paper synthesizes an operational governance evidence framework composed of: structural accountability collapse diagnostics, decision trace schemas, evidence sufficiency measurement, and label-free monitoring, integrated into a chain.
Methodological contribution: authors construct and present a synthesized framework from those four components (conceptual/analytical synthesis).
The Barcelona Declaration offers a promising forum for boundary governance.
Policy recommendation pointing to an existing initiative (Barcelona Declaration) as a suitable forum; stated without empirical evaluation in the excerpt.
Governance should calibrate the annulus, not abolish it: thin enough to serve research efficiently, wide enough to sustain innovation.
Normative policy recommendation from the authors; based on their conceptual framework rather than on empirical policy evaluation in the excerpt.
Artificial intelligence reshapes the annulus by lowering barriers to basic structuring.
Conceptual claim in the paper; asserted as an effect of AI on metadata production without empirical estimates in the excerpt.
States can adjust their foreign policies to this fact by focusing on resilience, technological sovereignty, strategic decoupling, and coordination through alliances.
Policy-prescriptive recommendations based on the paper's theoretical framework and analysis; no empirical testing or sample size reported in the abstract.
ClawNet enables multiple users to collaborate securely through their respective agents.
Capability claim about the instantiated system (authors assert that ClawNet enables secure multi-user collaboration; excerpt contains no empirical security evaluation or user study).
We instantiate this paradigm in ClawNet, an identity-governed agent collaboration framework that enforces identity binding and authorization verification through a central orchestrator.
Implementation claim: authors state they built ClawNet as an instantiation of their paradigm (paper describes framework/architecture; no experimental evaluation included in excerpt).
Action-level accountability logs every operation against its owner's identity and authorization, ensuring full auditability.
Design claim describing an accountability primitive (paper asserts logging and auditability as a property; no audit or verification evidence shown in excerpt).
Scoped authorization enforces per-identity access control and escalates boundary violations to the owner.
Design/specification claim describing the scoped authorization governance primitive in the proposed paradigm (no empirical or security evaluation provided in excerpt).