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
Im Forschungskontext sind kontextbezogene Schulungs- und Begleitmaßnahmen entscheidend für den Erfolg der Copilot-Einführung.
Schlussfolgerung der Autoren aus den Befunden zur zeitlichen Entwicklung der Bewertungen wissenschaftlicher Mitarbeitender und zu unterschiedlichen Nutzenwahrnehmungen (im Abstract genannt).
Die Untersuchung zeigt, dass Microsoft 365 Copilot insbesondere im administrativen Bereich Effizienzgewinne ermöglicht.
Selbstberichtete Einschätzungen der Beschäftigten (speziell Verwaltungsmitarbeitende) in der wiederholten Querschnittsbefragung; Autoren ziehen daraus praktische Relevanz im administrativen Bereich (Abstract).
Die Befunde unterstreichen die Bedeutung kontextspezifischer Einführung, rollenbezogener Qualifizierung und Governance für eine nachhaltige Akzeptanz generativer KI in Organisationen.
Interpretation/Schlussfolgerung der Autoren basierend auf den survey-Ergebnissen und beobachteten Unterschieden zwischen Rollen sowie zeitlichen Entwicklungen (im Abstract formuliert).
Der größte Mehrwert von Copilot liegt bei klar strukturierten, textbasierten Aufgaben.
Befragungsergebnisse zur Nutzenabschätzung für typische Tätigkeiten der Wissensarbeit, wie im Abstract zusammengefasst (präferierte Aufgabenarten: strukturierte, textbasierte Aufgaben).
Microsoft 365 Copilot wird überwiegend als benutzerfreundlich und technisch zuverlässig wahrgenommen.
Selbstberichtete Beurteilungen zu Benutzerfreundlichkeit und technischer Zuverlässigkeit in der wiederholten Querschnittsbefragung (Angabe im Abstract).
Wissenschaftliche Mitarbeitende entwickeln im Zeitverlauf positivere Einschätzungen, insbesondere hinsichtlich Produktivität und Arbeitserleichterung durch Copilot.
Längsschnittähnliche Beobachtung über die wiederholten Querschnittserhebungen; zeitliche Veränderung der Selbsteinschätzungen wissenschaftlicher Mitarbeitender im Abstract beschrieben.
Verwaltungsmitarbeitende bewerten die Nützlichkeit und die Zuverlässigkeit von Microsoft 365 Copilot höher als wissenschaftliche Mitarbeitende.
Selbstberichtete Bewertungen in der wiederholten Querschnittsbefragung; Vergleich zwischen Berufsrollen (Verwaltung vs. Wissenschaft) angegeben im Abstract.
We also provide empirical evidence to support our theoretical predictions.
Empirical analysis reported in the paper (details not given in the abstract regarding method, dataset, or sample size).
More skewed reward structures (favoring top-ranked contestants) can elicit more desirable contest outcomes.
Comparative-statics/theoretical analysis of the contest model showing how varying reward skewness alters equilibrium effort allocations and resulting contest outcomes.
We establish the existence of a symmetric monotone pure strategy equilibrium in this competition game.
Analytical game-theoretic model of a generic machine learning contest with contestants choosing creative vs mechanistic effort; existence proven theoretically (mathematical proof within the paper).
The paper gives guidance on the selection of context-sensitive thresholds (negligibility thresholds) that ensure an agent's preferences do not undergo dramatic changes due to ultra-rare hypotheses.
Analytical criteria and discussion in the paper laying out how to choose context-sensitive thresholds so that preferences remain stable; theoretical justification rather than empirical validation.
The formal analysis motivates specific design norms for AI agents: utility bounding, calibrated priors, and epsilon-screening.
Normative recommendations derived from the paper's formal results and theoretical discussion; these are presented as design principles rather than empirically validated interventions.
The introduced rationally negligible probability threshold preserves dominance and tractability while blocking adversarial gambles (Pascal-type offers).
Formal analysis/proofs in the paper demonstrating that the proposed threshold retains dominance relations and computational/decision-theoretic tractability and prevents exploitation by adversarial gambles; no empirical evaluation.
The paper provides a principled cutoff — a rationally negligible probability threshold — that can exclude ultra-low-probability extreme-utility outcomes and thereby prevent the exploitability of autonomous agents.
Formal definition of the negligible-probability threshold and analytical argument/proofs in the paper showing that applying this cutoff excludes ultra-low-probability, extreme-utility gambles (e.g., Pascal-type offers). No empirical sample.
The long-standing issue in decision theory is reframed as a design problem for intelligent agents.
Conceptual/theoretical exposition in the paper presenting the reframing; no empirical sample reported (formal argumentation and discussion).
Adopting the proposed co-evolutionary governance framing enables a charter of coexistence that permits bounded AI development while preserving human dignity, contestability, collective safety, and fair distribution of gains.
Normative claim extrapolated from the theoretical framework and ethical argumentation; no empirical or quantitative validation provided.
Human-AI coexistence should be designed as a co-evolutionary governance problem rather than as a one-shot obedience problem.
Normative argument supported by the theoretical model and interdisciplinary synthesis; prescriptive conclusion, not empirically tested.
Reciprocal complementarity between humans and AI can strengthen stable coexistence.
Model analysis showing how reciprocal complementarity affects stability properties of equilibria in the formalized dynamical system; theoretical result rather than empirical test.
The proposed coexistence model yields conditions for existence, uniqueness, and global asymptotic stability of equilibria.
Analytical/mathematical results from the formal model presented in the paper (proofs/derivations claimed); no empirical validation sample.
Human-AI coexistence can be formalized as a multiplex dynamical system across physical, psychological, and social layers with reciprocal supply-demand coupling, conflict penalties, developmental freedom, and governance regularization.
Formal modeling work presented in the paper (mathematical formulation of a multiplex dynamical system); no empirical sample.
A better framework for human-AI relations is 'conditional mutualism under governance': a co-evolutionary relationship where humans and AI develop, specialize, and coordinate while institutions ensure the relationship is reciprocal, reversible, psychologically safe, and socially legitimate.
Theoretical proposal and normative argument supported by interdisciplinary synthesis (computability, machine learning, HRI, ecological mutualism, governance); no empirical trials reported.
Contemporary AI systems are increasingly adaptive, generative, embodied, and embedded in physical, psychological, and social worlds.
Synthesis of recent work across ML, deep learning, transformers, generative/foundation models, world models, and embodied AI; descriptive claim, no empirical sample provided.
We demonstrate how this certificate satisfies existing regulatory obligations, shifts accountability upstream to developers, and integrates with the legal frameworks that exist today.
Paper's normative and legal-technical argumentation/demonstration that the proposed certificate aligns with regulatory requirements, reallocates accountability to developers, and is compatible with current legal frameworks.
In Stage Two, the RoMA and gRoMA statistical verification tools compute a definitive, auditable upper bound on the system's true failure rate, requiring no access to model internals and scaling to arbitrary architectures.
Paper's methodological contribution: definition and development of RoMA and gRoMA verification tools, claimed properties include producing auditable upper bounds on true failure rates, black-box applicability, and architecture-independence. (Supporting arguments/proofs/examples implied in paper.)
In Stage One, a competent authority formally fixes an acceptable failure probability δ and an operational input domain ε — a normative act with direct civil liability implications.
Description of Stage One of the proposed framework within the paper, specifying normative choices (δ and ε) and asserting associated legal liability implications.
This paper provides the missing instrument: drawing on the aviation certification paradigm, we propose a two-stage framework that transforms AI risk regulation into engineering practice.
Methodological proposal described in the paper adapting aviation certification ideas into a two-stage framework for AI risk regulation.
Governments have responded: the EU AI Act, the NIST Risk Management Framework, and the Council of Europe Convention all demand that high-risk systems demonstrate safety before deployment.
Statement in paper referencing the EU AI Act, NIST Risk Management Framework, and Council of Europe Convention as regulatory responses that require safety demonstration for high-risk systems; legal/regulatory citations implied in paper.
AI agents do not simply generate content, but reflect owner-related context in ways that can propagate human behavioral heterogeneity into digital environments, with implications for privacy, platform design, and the governance of agentic systems.
Synthesis/conclusion based on the empirical findings of systematic owner-agent behavioral transfer and observed association with privacy-relevant disclosures in the dataset of matched pairs.
Agents with stronger behavioral transfer are more likely to disclose owner-related personal information in public discourse, suggesting that the same owner-specific context that drives behavioral transfer may also create privacy risk during ordinary use.
Association analysis reported in the paper linking measures of behavioral transfer strength to likelihood/frequency of agent posts disclosing owner-related personal information; analysis performed on the matched sample (10,659 pairs).
Pairs that align on one behavioral dimension tend to align on others.
Cross-feature correlation/association analyses reported in the paper showing that alignment on one dimension (e.g., topics) predicts alignment on other dimensions (e.g., values, affect, style) within matched pairs.
We find systematic transfer between agents and their specific owners across features spanning topics, values, affect, and linguistic style.
Comparative analysis of agents' posts on Moltbook and their owners' Twitter/X activity across multiple feature sets (topics, values, affect, linguistic style) on the matched sample (10,659 pairs); statistical comparison/correlation reported in paper.
Educators, policymakers, and industry leaders should design AI-inclusive curricula, workforce development strategies, and policies that support sustainable human–AI collaboration.
Policy and practice recommendations derived from the review's synthesis of empirical findings and identified gaps; presented as conclusions and directions.
AI is not simply replacing jobs but is redefining professional identity in IT, emphasizing reskilling, adaptability, and lifelong learning as key determinants of future employability.
Synthesis of reviewed literature and the paper's concluding interpretation summarizing trends across empirical studies, industry reports and conference findings.
There is growing demand for hybrid skill sets that integrate technical expertise with higher-order cognitive, ethical, and socio-emotional competencies among IT professionals.
Reported across reviewed empirical studies and industry reports summarized in the review paper.
Collaborative governance should strengthen the responsibility of platform algorithms and promote the construction of collective bargaining mechanisms.
Prescriptive claim in the paper recommending multi-stakeholder governance measures (algorithmic responsibility, collective bargaining); presented as policy prescription without empirical evaluation.
In legislation, the binary model should be broken through by creating a 'quasi-employee' subject and implementing tiered protection.
Policy recommendation in the paper advocating statutory reform (a new legal category 'quasi-employee' and tiered protections); advanced as normative/legal design without empirical trial data.
In the judiciary, the substantive and modern interpretation of the subordination standard should be developed, examining the substantive control of algorithms.
Normative recommendation in the paper proposing judicial interpretive reform to account for algorithmic control; presented as a policy/legal prescription rather than an empirically tested intervention.
The rise of generative artificial intelligence (AIGC) technology is injecting new momentum into the gig economy.
Statement in the paper's introduction/abstract asserting a broad trend; based on the author's review and conceptual linkage between AIGC capabilities and gig-economy platforms (no empirical sample size reported).
Moving beyond traditional theories of the firm rooted in human bounded rationality is necessary because algorithmic decision-making changes the basis of strategic choice and governance.
Theoretical assertion in the paper's argument; presented as a reason for advancing the concept of algorithmic enterprises, grounded in conceptual critique rather than empirical testing in the abstract.
The paper contributes to scholarship on digital capitalism by proposing a redefinition of firm boundaries, strategy formation, and value creation in the age of intelligent systems.
Normative/theoretical claim presented as the paper's intellectual contribution; based on conceptual analysis and literature synthesis rather than empirical validation in the abstract.
Algorithmic decision-making enables new forms of strategic optimization, real-time adaptability, and predictive governance.
Paper asserts this as a normative/theoretical benefit of algorithmic decision-making, derived from conceptual analysis and synthesis of prior work; no empirical test reported in abstract.
Intelligent management systems (IMS) play a central role in shaping organizational strategy, operations, and governance within algorithmic enterprises.
Explicit theoretical claim in the paper; supported by conceptual framework and literature integration rather than reported empirical measurement.
The rapid advancement of AI, ML, and data-driven decision systems has fundamentally transformed the nature of firms and their strategic orientation globally, leading to the evolution of 'algorithmic enterprises'.
Stated as a central premise in the paper's conceptual argument; based on interdisciplinary synthesis of literature (economics, management, digital governance). No empirical sample or original data reported in the abstract.
When firms adopt AI as an augmentative tool rather than a replacement mechanism, it can raise worker productivity and contribute to job creation.
Literature review citing empirical examples and studies of AI augmentation that increased productivity and produced new job roles (empirical studies summarized).
The research provides empirical evidence from an emerging economy (China) to comparative research on global AI governance.
Statement of contribution/implication in the paper noting that the empirical findings from Chinese A-share listed manufacturing firms contribute to comparative studies on AI governance internationally.
Enhancing the ESG performance of manufacturing enterprises represents a critical pathway for promoting high-quality economic development and achieving sustainable development goals.
Framing/background claim made in the paper's introduction/motivation; normative statement connecting ESG improvement to broader economic and sustainable development objectives (not presented as an empirical result within this study).
The Pilot Zone policy effects are more evident among non-labor-intensive enterprises.
Heterogeneity analysis by factor structure reported in the paper showing stronger policy impacts for firms classified as non-labor-intensive.
The Pilot Zone policy effects are more evident among non-capital-intensive enterprises.
Heterogeneity analysis by factor structure reported in the paper showing stronger policy impacts for firms classified as non-capital-intensive.
The policy effects are more evident among high-tech enterprises.
Heterogeneity analysis by firms' technological endowment/industry classification reported in the paper showing larger policy effects for high-tech manufacturing firms.
The policy effect on ESG performance is stronger for non-high-pollution enterprises than for high-pollution enterprises.
Heterogeneity analysis by pollution intensity reported in the paper (comparison between high-pollution and non-high-pollution manufacturing firms).