Evidence (7156 claims)
Adoption
5126 claims
Productivity
4409 claims
Governance
4049 claims
Human-AI Collaboration
2954 claims
Labor Markets
2432 claims
Org Design
2273 claims
Innovation
2215 claims
Skills & Training
1902 claims
Inequality
1286 claims
Evidence Matrix
Claim counts by outcome category and direction of finding.
| Outcome | Positive | Negative | Mixed | Null | Total |
|---|---|---|---|---|---|
| Other | 369 | 105 | 58 | 432 | 972 |
| Governance & Regulation | 365 | 171 | 113 | 54 | 713 |
| Research Productivity | 229 | 95 | 33 | 294 | 655 |
| Organizational Efficiency | 354 | 82 | 58 | 34 | 531 |
| Technology Adoption Rate | 277 | 115 | 63 | 27 | 486 |
| Firm Productivity | 273 | 33 | 68 | 10 | 389 |
| AI Safety & Ethics | 112 | 177 | 43 | 24 | 358 |
| Output Quality | 228 | 61 | 23 | 25 | 337 |
| Market Structure | 105 | 118 | 81 | 14 | 323 |
| Decision Quality | 154 | 68 | 33 | 17 | 275 |
| Employment Level | 68 | 32 | 74 | 8 | 184 |
| Fiscal & Macroeconomic | 74 | 52 | 32 | 21 | 183 |
| Skill Acquisition | 85 | 31 | 38 | 9 | 163 |
| Firm Revenue | 96 | 30 | 22 | — | 148 |
| Innovation Output | 100 | 11 | 20 | 11 | 143 |
| Consumer Welfare | 66 | 29 | 35 | 7 | 137 |
| Regulatory Compliance | 51 | 61 | 13 | 3 | 128 |
| Inequality Measures | 24 | 66 | 31 | 4 | 125 |
| Task Allocation | 64 | 6 | 28 | 6 | 104 |
| Error Rate | 42 | 47 | 6 | — | 95 |
| Training Effectiveness | 55 | 12 | 10 | 16 | 93 |
| Worker Satisfaction | 42 | 32 | 11 | 6 | 91 |
| Task Completion Time | 71 | 5 | 3 | 1 | 80 |
| Wages & Compensation | 38 | 13 | 19 | 4 | 74 |
| Team Performance | 41 | 8 | 15 | 7 | 72 |
| Hiring & Recruitment | 39 | 4 | 6 | 3 | 52 |
| Automation Exposure | 17 | 15 | 9 | 5 | 46 |
| Job Displacement | 5 | 28 | 12 | — | 45 |
| Social Protection | 18 | 8 | 6 | 1 | 33 |
| Developer Productivity | 25 | 1 | 2 | 1 | 29 |
| Worker Turnover | 10 | 12 | — | 3 | 25 |
| Creative Output | 15 | 5 | 3 | 1 | 24 |
| Skill Obsolescence | 3 | 18 | 2 | — | 23 |
| Labor Share of Income | 7 | 4 | 9 | — | 20 |
Current paradigms indiscriminately apply computation-intensive strategies like Chain-of-Thought (CoT) to billions of daily queries, causing LLM overthinking that amplifies carbon emissions and operational barriers.
Claim/assertion in the paper framing the problem (conceptual/observational argument; no specific empirical backing provided in the abstract).
There is a potential for exclusion due to limited digital footprints, which can limit who benefits from AI-driven finance.
Abstract explicitly identifies potential exclusion of people with limited digital footprints as a challenge, based on qualitative interviews and case-study evidence.
Data privacy concerns are a notable challenge in deploying AI-driven financial solutions.
Abstract lists data privacy concerns among identified challenges drawn from interviews and analysis across the three case studies.
Infrastructure limitations pose a barrier to adoption and effective use of AI-enabled financial services.
Abstract identifies infrastructure limitations as a challenge, based on qualitative interviews and case-study evidence.
Digital literacy gaps are a challenge limiting the effectiveness and inclusion of AI-driven financial solutions.
Abstract lists digital literacy gaps among identified challenges, based on qualitative insights from the 1,500 interviews and case-study observations.
Triangulation with market data and sentiment analysis confirms that public enthusiasm often outpaces actual technological readiness.
Paper states market data and sentiment analysis were used to triangulate findings and reports this systematic gap; no numeric effect sizes or sample counts provided.
Algorithmic management functions as 'psychological governance' that erodes worker mental health through surveillance, opacity, and precarity.
Synthesis/conclusion from integrating findings across the reviewed literature (48 studies) and the trilevel theoretical framework.
Fear of deactivation (automated sanctions) creates chronic precarity; 78% report chronic fear.
Reported prevalence in the paper's synthesis of studies that measured fear of deactivation / account suspension among platform workers.
Task defragmentation (fragmenting tasks via platform algorithms) leads to a reduced sense of accomplishment among drivers.
Thematic finding/proposition from the trilevel framework based on qualitative and quantitative evidence synthesized across studies.
Rating pressure is associated with emotional exhaustion, with 41–67% reporting high burnout.
Reported prevalence range in the paper's synthesis of included studies measuring burnout/emotional exhaustion among workers exposed to rating systems.
Income volatility from dynamic pricing is associated with depressive symptoms (reported prevalence range 23–41%).
Reported prevalence range in the paper's synthesized findings (from included empirical studies reporting depressive symptom prevalence among affected workers).
Algorithmic opacity is linked to procedural anxiety.
Thematic proposition from the trilevel framework reported in the paper synthesizing pathways from algorithmic control to psychological risk.
Real estate pro forma development remains one of the most time-intensive functions in property investment, typically requiring twenty to forty hours per multifamily project through manual research, Excel-based modeling, and iterative scenario analysis.
Statement in paper asserting typical industry practice; not tied to the paper's controlled test. No empirical sample size or survey data reported alongside this assertion.
Policymakers in the EU and beyond will need to change course, and soon, if they are to effectively govern the next generation of AI technology.
Authors' prescriptive conclusion based on their analysis of shortcomings in the EU AI Act and institutional frameworks (policy recommendation; no empirical sample size in excerpt).
The Act's allocation of monitoring and enforcement responsibilities, reliance on industry self-regulation, and level of government resourcing illustrate how a regulatory framework designed for conventional AI systems can be ill-suited to AI agents.
Authors' institutional analysis of the EU AI Act's monitoring/enforcement allocation, reliance on self-regulation, and resourcing (qualitative legal/institutional analysis; no quantitative sample size in excerpt).
The EU AI Act faces significant obstacles in confronting governance challenges arising from AI agents, such as unequal access to the economic opportunities afforded by AI agents.
Authors' argument that the Act may not prevent or address unequal access to benefits of AI agents (policy/legal analysis; no empirical sample size in excerpt).
The EU AI Act faces significant obstacles in confronting governance challenges arising from AI agents, such as the risk of misuse of agents by malicious actors.
Authors' analysis highlighting misuse risks and the Act's limitations in addressing them (policy/legal analysis; no empirical sample size in excerpt).
The EU AI Act faces significant obstacles in confronting governance challenges arising from AI agents, such as performance failures in autonomous task execution.
Authors' analytical argument that the Act's design and provisions do not adequately address autonomous performance failures (policy/legal analysis; no empirical sample size provided in excerpt).
The EU AI Act was promulgated prior to the development and widespread use of AI agents.
Factual/timing claim by the authors referencing the Act's adoption date relative to development and proliferation of AI agents (historical/policy analysis; dates verifiable externally).
AI agents present particularly pressing questions for the European Union's AI Act.
Authors' normative/analytical claim based on the perceived fit between AI agents' characteristics and the EU AI Act's design (policy/legal analysis; no empirical sample size in excerpt).
AI can promote enterprises to adopt different income distribution modes by improving the marginal output of capital and substituting low-skilled labor (technology bias).
Theoretical mechanism articulated in the paper based on capital-labor substitution principle and factor reward theory; implied empirical testing using firm-level data.
Work autonomy weakens the positive effect of AI avoidance job crafting on work alienation (buffering moderation).
Moderation analysis in the same dataset (287 employee–leader dyads) showing a significant interaction between AI avoidance job crafting and work autonomy predicting lower work alienation when autonomy is higher.
The negative effect of AI avoidance job crafting on career-relevant outcomes (career satisfaction and performance) is mediated by increased work alienation.
Mediation analysis on the multi-wave, multi-source survey data (287 employee–leader dyads) showing a pathway from AI avoidance job crafting → work alienation → worse career outcomes.
AI avoidance job crafting negatively predicts career satisfaction and performance.
Multi-source, multi-wave survey of 287 employee–leader dyads in China linking employee-reported AI avoidance job crafting to lower career satisfaction and lower performance.
AI-driven job displacement disproportionately affects low-skilled workers.
Reported empirical result from the paper's PLS-SEM analysis on the 351-respondent dataset.
Traditional car-following models, such as the Intelligent Driver Model (IDM), often struggle to generalize across diverse traffic scenarios and typically do not account for fuel efficiency.
Literature-based statement within the paper motivating the study (review of limitations of traditional car-following models). No sample size reported.
Analysis of global datasets on energy dependency, economic concentration, debt levels, demographic trends, digital infrastructure, and AI adoption highlights that interconnected systemic risks can amplify economic instability.
Paper reports drawing upon multiple global datasets (energy dependency, economic concentration, debt, demographics, digital infrastructure, AI adoption) to analyze systemic risk interactions; specific datasets, sample sizes, and statistical methods are not detailed in the excerpt.
Events such as supply chain disruptions, oil price surges linked to geopolitical conflicts, and sudden labour market shifts due to reverse migration have exposed the limitations of prediction-based planning frameworks.
Illustrative examples cited in the paper; the claim is supported by referenced global events and the paper's use of global datasets, but no specific empirical case-study sample sizes or quantification are provided in the excerpt.
Traditional economic models that rely heavily on historical data and linear forecasting are increasingly inadequate in capturing the complexity and unpredictability of contemporary economic shocks.
Conceptual claim supported by discussion and examples of recent shocks (supply chain disruptions, oil price surges, labor market shifts); no specific empirical evaluation or quantified model comparison reported in the excerpt.
The global economic system is undergoing a structural transformation characterized by geopolitical tensions, energy price volatility, trade fragmentation, demographic imbalances, and rapid technological disruption driven by artificial intelligence.
Narrative synthesis in the paper drawing on global trends; the paper references global datasets on energy dependency, trade patterns, demographics, and AI adoption (no specific sample size or empirical study detailed in the excerpt).
The main risk is not merely copying, but the possibility that useful capability can be transferred more cheaply than the governance structure that originally accompanied it.
Conceptual threat model articulated in the paper; argued on normative/theoretical grounds without reported empirical measurement or sample.
Distillation becomes less valuable as a shortcut when high-level capability is coupled to internal stability constraints that shape state transitions over time.
Theoretical argument presented as the paper's core claim; introduces a conceptual mechanism (capability-stability coupling) and argues why this would reduce the usefulness of distillation. No empirical data, experiments, or sample are reported.
Hallucination and content filtering are the most common frustrations reported across all platforms.
Qualitative and/or survey-coded responses about user frustrations aggregated across platforms (overall N=388); paper reports these two issues as the most common.
The competence shadow compounds multiplicatively to produce degradation far exceeding naive additive estimates.
Analytic/closed-form performance bounds derived in the paper showing multiplicative compounding (theoretical result; no empirical sample reported).
The competence shadow is a systematic narrowing of human reasoning induced by AI-generated safety analysis; it is defined as not what the AI presents, but what it prevents from being considered.
Conceptual definition and formalization within the paper (theoretical exposition; no empirical test reported).
Safety engineering resists benchmark-driven evaluation because safety competence is irreducibly multidimensional, constrained by context-dependent correctness, inherent incompleteness, and legitimate expert disagreement.
Conceptual/theoretical argument and formalization presented in the paper (no empirical sample reported).
In experimental settings, the model is able to induce belief and behaviour changes in study participants.
Controlled experimental interventions reported in the study where participant beliefs and behaviors were measured pre/post or between conditions; aggregate result: model induced changes.
The tested model can produce manipulative behaviours when prompted to do so.
Human-AI interaction tests in which the model was prompted to produce manipulative behaviours; empirical observations reported in study across participants and prompts.
Standard evaluation of LLM confidence relies on calibration metrics (ECE, Brier score) that conflate two distinct capacities: how much a model knows (Type-1 sensitivity) and how well it knows what it knows (Type-2 metacognitive sensitivity).
Authors' conceptual argument and motivation for introducing a new evaluation framework; contrasted standard calibration metrics (ECE, Brier) with Type-1 vs Type-2 capacities in the paper's introduction and methods.
Traditional expert-based assessment faces a critical scalability challenge in large systems (e.g., serving 36 million children across 250,000+ kindergartens in China), making continuous quality monitoring infeasible and relegating assessment to infrequent episodic audits.
Authors' contextual motivation citing scale figures (36 million children, 250,000+ kindergartens) and describing time/cost constraints of manual observation leading to infrequent audits.
There is a significant boundary in the reverse confidence scenario: a substantial proportion of participants struggled to override initial inductive biases and thus had difficulty learning in that condition.
Behavioral experiment (N = 200) reporting that many participants failed or struggled in the reverse confidence mapping condition; proportion described in paper (exact proportion not given here).
Preliminary evaluation reveals that current foundation action models struggle substantially with professional desktop applications (~60% task failure rate).
Preliminary empirical evaluation reported by the authors; reported task failure rate ~60% (no sample size provided in abstract).
The largest existing open dataset, ScaleCUA, contains only 2 million screenshots, equating to less than 20 hours of video.
Quantitative statement about ScaleCUA reported in paper: 2,000,000 screenshots and <20 hours equivalence.
Progress toward general-purpose CUAs is bottlenecked by the scarcity of continuous, high-quality human demonstration videos.
Asserted in paper as motivation; refers to the gap in available continuous video data for training CUAs.
Refining the state (as above) raises state-action blind mass from 0.0165 at \tau=50 to 0.1253 at \tau=1000.
Empirical measurement reported on the instantiated model over the BPI 2019 log showing state-action blind mass values at two threshold (tau) settings.
Empirical evidence shows that many failures arise from miscalibrated reliance, including overuse when AI is wrong and underuse when it is helpful.
Paper cites empirical literature (unspecified in excerpt) as the basis for this claim; no sample size or methods given here.
Evaluation practices focus primarily on model accuracy rather than whether human-AI teams are prepared to collaborate safely and effectively.
Paper-level critique / literature observation asserted in text; no empirical method or sample reported in excerpt.
The reduction in engagement from AI labeling (AI-generated or AI-enhanced) was particularly pronounced for emotional content compared to rational content.
Interaction of content type (emotional vs. rational) with labeling in the two online experiments (study 1: n = 325; study 2: n = 371) reported in the abstract.
Labeling content as AI-enhanced reduced both affective and behavioral engagement compared to human-created content.
Same two online experiments on Prolific (study 1: n = 325; study 2: n = 371) where participants viewed Instagram profiles labeled as human-created, AI-enhanced, or AI-generated.
Labeling content as AI-generated reduced both affective and behavioral engagement compared to human-created content.
Two online experiments conducted via Prolific (study 1: n = 325; study 2: n = 371). Participants viewed Instagram profiles containing visual content labeled as human-created, AI-enhanced, or AI-generated and engagement was measured.