|
on Artificial Intelligence |
| By: | Jason Hartline |
| Abstract: | A fundamental challenge for modern economics is to understand what happens when actors in an economy are replaced with algorithms. Like rationality has enabled understanding of outcomes of classical economic actors, no-regret can enable the understanding of outcomes of algorithmic actors. This review article covers the classical computer science literature on no-regret algorithms to provide a foundation for an overview of the latest economics research on no-regret algorithms, focusing on the emerging topics of manipulation, statistical inference, and algorithmic collusion. |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2601.22079 |
| By: | Joshua S. Gans |
| Abstract: | Machine learning systems embed preferences either in training losses or through post-processing of calibrated predictions. Applying information design methods from Strack and Yang (2024), this paper provides decision problem agnostic conditions under which separation training preference free and applying preferences ex post is optimal. Unlike prior work that requires specifying downstream objectives, the welfare results here apply uniformly across decision problems. The key primitive is a diminishing-value-of-information condition: relative to a fixed (normalised) preference-free loss, preference embedding makes informativeness less valuable at the margin, inducing a mean-preserving contraction of learned posteriors. Because the value of information is convex in beliefs, preference-free training weakly dominates for any expected utility decision problem. This provides theoretical foundations for modular AI pipelines that learn calibrated probabilities and implement asymmetric costs through downstream decision rules. However, separation requires users to implement optimal decision rules. When cognitive constraints bind, as documented in human AI decision-making, preference embedding can dominate by automating threshold computation. These results provide design guidance: preserve optionality through post-processing when objectives may shift; embed preferences when decision-stage frictions dominate. |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2601.18732 |
| By: | Wayne Gao; Sukjin Han; Annie Liang |
| Abstract: | Large language models (LLMs) are increasingly used to predict human behavior. We propose a measure for evaluating how much knowledge a pretrained LLM brings to such a prediction: its equivalent sample size, defined as the amount of task-specific data needed to match the predictive accuracy of the LLM. We estimate this measure by comparing the prediction error of a fixed LLM in a given domain to that of flexible machine learning models trained on increasing samples of domain-specific data. We further provide a statistical inference procedure by developing a new asymptotic theory for cross-validated prediction error. Finally, we apply this method to the Panel Study of Income Dynamics. We find that LLMs encode considerable predictive information for some economic variables but much less for others, suggesting that their value as substitutes for domain-specific data differs markedly across settings. |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2601.12343 |
| By: | Pietro Bini; Lin William Cong; Xing Huang; Lawrence J. Jin |
| Abstract: | Do generative AI models, particularly large language models (LLMs), exhibit systematic behavioral biases in economic and financial decisions? If so, how can these biases be mitigated? Drawing on the cognitive psychology and experimental economics literatures, we conduct the most comprehensive set of experiments to date—originally designed to document human biases—on prominent LLM families across model versions and scales. We document systematic patterns in LLM behavior. In preference-based tasks, responses become more human-like as models become more advanced or larger, while in belief-based tasks, advanced large-scale models frequently generate rational responses. Prompting LLMs to make rational decisions reduces biases. |
| JEL: | D03 G02 G11 G4 G40 G41 |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:34745 |
| By: | Simon Gaechter (University of Nottingham); Dominik Suri (University of Bonn); Sebastian Kube (University of Bonn) |
| Abstract: | AI-driven systems are rapidly moving from decision support to directing human behavior through rules, recommendations, and compliance requests. This shift expands everyday human–AI interaction and raises the possibility that AI may function as an authority figure. However, the behavioral consequences of AI as an authority figure remain poorly understood. We investigate whether individuals differ in their willingness to comply with arbitrary rules depending on whether these rules are attributed to an AI agent (ChatGPT) or to a fellow human. In a between-subject design, 977 US Prolific users completed the coins task: they could earn a monetary payoff by stopping the disappearance of coins at any time, but a rule instructed them to wait for a signal before doing so. There are no conventional reasons to follow this rule: complying is costly and nobody is harmed by non-compliance. Despite this, we find high rule-following rates: 64.3% followed the rule set by ChatGPT and 63.9% complied with the human-set rule. Descriptive and normative beliefs about rule following, aswell as compliance conditional on these beliefs, are also largely unaffected by the rule’s origin. However, subjective social closeness to the rule setter significantly predicts how participants condition their behavior on social expectations: when participants perceive the rule setter as subjectively closer, conditional compliance is higher and associated beliefs are stronger, irrespective of whether the rule setter is human or AI. |
| Keywords: | artificial intelligence; AI-human interaction; ChatGPT; rule-following; coins task; CRISP framework; social expectations; conditional rule conformity; social closeness; IOS11; online experiments |
| Date: | 2026–02 |
| URL: | https://d.repec.org/n?u=RePEc:not:notcdx:2026-02 |
| By: | J. Ignacio Conde-Ruiz; Miguel Díaz Salazar; Juan José Ganuza |
| Abstract: | This paper combines artificial intelligence with economic modeling to design evaluation committees that are both efficient and fair in the presence of gender differences in economic research orientation. We develop a dynamic framework in which research evaluation depends on the thematic similarity between evaluators and researchers. The model shows that while topic balanced committees maximize welfare, this researchneutral-gender allocation is dynamically unstable, leading to the persistent dominance of the group initially overrepresented in evaluation committees. Guided by these predictions, we employ unsupervised machine learning to extract research profiles for male and female researchers from articles published in leading economics journals between 2000 and 2025. We characterize optimal balanced committees within this multidimensional latent topic space and introduce the Gender-Topic Alignment Index (GTAI) to measure the alignment between committee expertise and female-prevalent research areas. Our simulations demonstrate that AI-based committee designs closely approximate the welfare-maximizing benchmark. In contrast, traditional headcount-based quotas often fail to achieve balance and may even disadvantage the groups they intend to support. We conclude that AI-based tools can significantly optimize institutional design for editorial boards, tenure committees, and grant panels. |
| Keywords: | machine learning, artificial intelligence, Topic Modeling, evaluation committees, committee quotas, research orientation |
| JEL: | D72 D82 J16 J78 |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:upf:upfgen:1937 |
| By: | Sueyoul Kim; Ginger Zhe Jin; Eungik Lee |
| Abstract: | Using a comprehensive dataset of posts from a major platform for anime- and manga-style artwork, we study the impact of the launch of a prominent text-to-image generative AI. Focusing on the majority of incumbent creators who do not adopt AI as a primary tool, we show that the AI launch led to a significant decline in post uploads by illustrators, whereas comic artists were less affected, reflecting the need for tight stylistic alignment across sequential images in comics. We present empirical evidence for two underlying mechanisms. First, illustration posts experience a loss of viewer attention, measured by bookmarks, following the AI launch, which can significantly harm creators’ business models. Second, direct competition from AI-generated content plays an important role: illustrators working on intellectual properties (IPs, such as Pokémon) that are more heavily invaded by AI reduce their uploads disproportionately more. We further examine creators’ responses and show that illustrators with greater exposure to AI avoid using tags favored by AI-generated content after the AI launch and broaden the range of IPs they work on, consistent with a risk-hedging response to AI invasion. |
| JEL: | D22 J24 L86 O14 O33 |
| Date: | 2026–01 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:34733 |
| By: | Ahmet Koseoglu (Erciyes University, Kayseri, Turkiye); Ali Gokhan Yucel (Erciyes University, Kayseri, Turkiye) |
| Abstract: | This study investigates the impact of artificial intelligence (AI) on employment in a panel of selected countries. Using a dynamic framework, we employ the two-step System Generalized Method of Moments (System-GMM) estimator with Windmeijer correction to address endogeneity and account for the persistence of labor market dynamics. High-quality AI publications are used as a proxy to measure AI development. Employment is disaggregated by gender, skill level, and age groups to capture heterogeneous effects across the labor force. The empirical results indicate that AI adoption exerts differentiated effects on employment, with younger and low-skilled workers being more exposed to displacement risks, while high-skilled groups show signs of complementarity. These findings suggest that the labor market implications of AI are uneven and depend on demographic and skill characteristics. Policy implications emphasize the importance of targeted education, skill upgrading, and adaptive labor market policies to mitigate risks and harness the potential benefits of AI-driven technological change. |
| Keywords: | Artificial Intelligence, employment, system GMM |
| Date: | 2025–08 |
| URL: | https://d.repec.org/n?u=RePEc:smo:raiswp:0552 |
| By: | Aldasoro, Iñaki; Gambacorta, Leonardo; Pal, Rozalia; Revoltella, Debora; Weiss, Christoph; Wolski, Marcin |
| Abstract: | This paper provides new evidence on how the adoption of artificial intelligence (AI) affects productivity and employment in Europe. Using matched EIBIS-ORBIS data on more than 12, 000 non-financial firms in the European Union (EU) and United States (US), we instrument the adoption of AI by EU firms by assigning the adoption rates of US peers to isolate exogenous technological exposure. Our results show that AI adoption increases the level of labor productivity by 4%. Productivity gains are due to capital deepening, as we find no adverse effects on firm-level employment. This suggests that AI increases worker output rather than replacing labor in the short run, though longer-term effects remain uncertain. However, productivity benefits of AI adoption are unevenly distributed and concentrate in medium and large firms. Moreover, AI-adopting firms are more innovative and their workers earn higher wages. Our analysis also highlights the critical role of complementary investments in software and data or workforce training to fully unlock the productivity gains of AI adoption. |
| Keywords: | Artificial intelligence, firm productivity, Europe, digital transformation |
| JEL: | D22 J24 L25 O33 O47 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:eibwps:335876 |
| By: | Piotr Lewandowski (Institute for Structural Research); Karol Madoń (Institute for Structural Research); Albert Park (Asian Development Bank) |
| Abstract: | This paper develops a task-adjusted, country-specific measure of workers’ exposure to artificial intelligence (AI) across 108 countries. Building on Felten et al. (2021), we adapt the artificial intelligence occupational exposure (AIOE) index to worker-level data from the Programme for the International Assessment of Adult Competencies (PIAAC) and extend it globally using comparable surveys and regression-based predictions, covering about 89% of global employment. Accounting for country-specific task structures reveals substantial cross-country heterogeneity: workers in low-income countries exhibit AI exposure levels roughly 0.8 United States (US) standard deviations below those in high-income countries, largely due to differences in within-occupation task content. Regression decompositions attribute most cross-country variation to information and communications technology intensity and human capital. High-income countries employ the majority of workers in highly AI-exposed occupations, while low-income countries concentrate in less exposed ones. Using two PIAAC cycles, we document rising AI exposure in high-income countries, driven by shifts in within-occupation tasks rather than employment structure. |
| Keywords: | job tasks;occupations;AI;technology;skills |
| JEL: | J21 J23 J24 |
| Date: | 2026–01–30 |
| URL: | https://d.repec.org/n?u=RePEc:ris:adbewp:022156 |
| By: | Kauhanen, Antti; Rouvinen, Petri |
| Abstract: | Abstract We examine the impact of generative AI on the youth labor market in Finland by replicating the key analyses of Brynjolfsson et al. (2025) with comprehensive population-level data. Contrary to the US findings, we find no systematic displacement effects linked to AI exposure among youth in Finland. Employment trends reflect demographic shifts rather than AI-driven changes, with early career groups showing modest declines and senior workers experiencing growth. Wage trajectories show no persistent differences across AI exposure levels. These results suggest that Finland’s labor market is resilient to immediate AI-induced disruptions in entry-level roles, likely because of structural and policy factors. |
| Keywords: | Generative artificial intelligence, Technological change, Employment, Wages, Occupations |
| JEL: | E24 J21 O33 |
| Date: | 2026–01–27 |
| URL: | https://d.repec.org/n?u=RePEc:rif:wpaper:135 |
| By: | Alex Haag |
| Abstract: | Global competition in artificial intelligence (AI) has intensified in recent years. Some assessments emphasize US exceptionalism, while others argue that China is eroding US dominance. By contrast, the progress of other advanced foreign economies (AFEs) receives far less attention. |
| Date: | 2025–10–06 |
| URL: | https://d.repec.org/n?u=RePEc:fip:fedgfn:102359 |
| By: | Mijalche Santa (Faculty of Economics-Skopje, Ss. Cyril and Methodius University in Skopje, North Macedonia); Blerton Zejneli (Faculty of Economics-Skopje, Ss. Cyril and Methodius University in Skopje, North Macedonia) |
| Abstract: | This research explores how social capital supports the adoption of artificial intelligence (AI) in developing countries, focusing on the role of "social brokers." A social broker is a trusted individual who occupies a unique position within a network, connecting individuals from different networks or maintaining connections with a larger number of individuals within the existing network. Based on input from the initial phase of the project, conducted in a developing country with high internet use but low AI adoption, we use qualitative research methods to better understand the practical aspects of AI adoption. Our early findings suggest that AI adoption goes beyond the right technology or skills and is strongly influenced by trusted communities and networks that shape decisions about AI adoption. "Social brokers" play a key role in this process. They help close knowledge gaps, address concerns of people who have not adopted AI or have adopted it at a low level, and show how AI can be relevant and useful for specific jobs and tasks. These "social brokers" are often seen as trusted friends, technology influencers, former colleagues, or respected local industry experts. Their presence and activities in tightly connected social networks appear to be very important for reducing the gap in AI adoption. The next phase of this research will focus on identifying the aspects of social capital that influence AI adoption, understanding the relationships that help overcome resistance to adopting AI, and developing strategies that use social capital to encourage faster AI adoption in developing countries. |
| Keywords: | AI, Technology adoption, Social brokers, Developing countries |
| JEL: | O31 O32 O33 |
| Date: | 2025–12–15 |
| URL: | https://d.repec.org/n?u=RePEc:aoh:conpro:2025:i:6:p:342-347 |
| By: | Rachid Azzaz (l'École Nationale Supérieure de Statistique et d'Économie Appliquée (ENSSEA)); Lylia Sami (Laboratory for Studies and Research in Digital Economy (LEREN)) |
| Abstract: | Artificial intelligence (AI) is transforming global labor markets, offering opportunities to boost productivity and create industries while raising concerns about job displacement and inequality. For Algeria, an oil-dependent economy, AI presents opportunities to diversify and improve efficiency as well as risks such as unemployment, skill gaps, and delayed adoption due to technological gaps and institutional constraints. This study adopts a novel approach that estimates automation risk by mixing the probabilities of the capabilities required for various occupations using occupational databases and crosswalks. These probabilities were adjusted with a correction factor accounting for the slower technology adoption in emerging markets, inspired by historical patterns. The findings reveal a significant lag in AI adoption, with Algeria’s automation trailing that of advanced economies by approximately 2.5 times the required time. Some qualitative insights from interviews with managers and employees are consistent with our results, and the study concludes that Algeria faces minimal immediate AI risks. Integration and its consequences are likely to be delayed due to industrial dependency and competitive pricing from developed countries. These findings provide a foundation for future MENA-wide studies on the impact of AI on labor markets. |
| Date: | 2025–10–20 |
| URL: | https://d.repec.org/n?u=RePEc:erg:wpaper:1796 |
| By: | Ivan Dionisijev (Faculty of Economics-Skopje, Ss. Cyril and Methodius University in Skopje, North Macedonia); Zorica Bozhinovska Lazarevska (Faculty of Economics-Skopje, Ss. Cyril and Methodius University in Skopje, North Macedonia); Todor Tocev (Faculty of Economics-Skopje, Ss. Cyril and Methodius University in Skopje, North Macedonia) |
| Abstract: | This study examines the integration of artificial intelligence (AI) in public sector internal auditing, focusing on the extent of AI adoption, the types of AI tools used, and the challenges faced by auditors in implementation. The research employs both descriptive statistical analysis and inferential techniques, including Spearman’s correlation and ANOVA, to assess the relationships between AI adoption and institutional factors. Findings indicate that while data analytics is the most commonly used AI tool, a significant proportion of respondents do not utilize AI in their auditing practices. The primary barriers to AI adoption include a lack of training, high costs, and concerns regarding data privacy. The study further reveals that AI usage varies depending on the type of institution in which they work. These insights contribute to the ongoing discussion on digital transformation in auditing, emphasizing the need for enhanced training programs and strategic investments to facilitate AI integration. |
| Keywords: | Internal auditing, Accounting information, Artificial Intelligence (AI), Public sector |
| JEL: | M42 H83 |
| Date: | 2025–12–15 |
| URL: | https://d.repec.org/n?u=RePEc:aoh:conpro:2025:i:6:p:22-45 |
| By: | Griesshaber, Niclas; Streb, Jochen |
| Abstract: | We leverage multimodal large language models (LLMs) to construct a dataset of 306, 070 German patents (1877-1918) from 9, 562 archival image scans using our LLM-based pipeline powered by Gemini-2.5-Pro and Gemini-2.5-Flash-Lite. Our benchmarking exercise provides tentative evidence that multimodal LLMs can create higher quality datasets than our research assistants, while also being more than 795 times faster and 205 times cheaper in constructing the patent dataset from our image corpus. About 20 to 50 patent entries are embedded on each page, arranged in a double-column format and printed in Gothic and Roman fonts. The font and layout complexity of our primary source material suggests to us that multimodal LLMs are a paradigm shift in how datasets are constructed in economic history. We open-source our benchmarking and patent datasets as well as our LLM-based data pipeline, which can be easily adapted to other image corpora using LLM-assisted coding tools, lowering the barriers for less technical researchers. Finally, we explain the economics of deploying LLMs for historical dataset construction and conclude by speculating on the potential implications for the field of economic history. |
| Keywords: | Multimodal Large Language Models, Information Extraction, Dataset Construction, German Patents |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:zbw:safewp:335887 |