|
on Artificial Intelligence |
| By: | Polachek, Solomon (Binghamton University, New York); Romano, Kenneth (State University of New York at Binghamton (Binghamton University)); Tonguc, Ozlem (State University of New York at Binghamton (Binghamton University)) |
| Abstract: | This study examines how large language models (LLMs) respond to varying stake sizes in the Dictator and Ultimatum games using the high-stakes design introduced by Andersen et al. (2011). We test ten leading LLMs chosen for their accessibility, prominence, and differences in reasoning capabilities. Results reveal substantial variation across models: Only 5 of 10 models exhibit strategic behavior by offering more in the Ultimatum Game (UG) than in the Dictator Game (DG). Relative to humans, 4 models are consistently more generous, 2 consistently less, and 4 vary with stake size. Only 1 model shows a monotonic decline in UG offers as stakes increase; the remaining 9 are non-monotonic or stable. Unlike humans, most models reduce UG offers when endowed with wealth. Prompting for “human-like†decisions generally increases generosity in the UG. These findings are important for evaluating whether LLMs can serve as realistic proxies for human subjects in behavioral experiments and highlight key limitations and future directions for model development. |
| Keywords: | ultimatum game, dictator game, fairness, payoff stakes, artificial intelligence |
| JEL: | D01 C72 C90 |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:iza:izadps:dp18545 |
| By: | Shuhuai Zhang; Shu Wang; Zijun Yao; Chuanhao Li; Xiaozhi Wang; Songfa Zhong; Tracy Xiao Liu |
| Abstract: | Altruism is fundamental to human societies, fostering cooperation and social cohesion. Recent studies suggest that large language models (LLMs) can display human-like prosocial behavior, but the internal computations that produce such behavior remain poorly understood. We investigate the mechanisms underlying LLM altruism using sparse autoencoders (SAEs). In a standard Dictator Game, minimal-pair prompts that differ only in social stance (generous versus selfish) induce large, economically meaningful shifts in allocations. Leveraging this contrast, we identify a set of SAE features (0.024% of all features across the model's layers) whose activations are strongly associated with the behavioral shift. To interpret these features, we use benchmark tasks motivated by dual-process theories to classify a subset as primarily heuristic (System 1) or primarily deliberative (System 2). Causal interventions validate their functional role: activation patching and continuous steering of this feature direction reliably shift allocation distributions, with System 2 features exerting a more proximal influence on the model's final output than System 1 features. The same steering direction generalizes across multiple social-preference games. Together, these results enhance our understanding of artificial cognition by translating altruistic behaviors into identifiable network states and provide a framework for aligning LLM behavior with human values, thereby informing more transparent and value-aligned deployment. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.19260 |
| By: | Irlenbusch, Bernd (University of Cologne); Rau, Holger (University of Duisburg-Essen and University of Göttingen); Rilke, Rainer (Economics Group, WHU – Otto Beisheim School of Management) |
| Abstract: | We study how human versus LLM-based evaluation and gender transparency shape entry into competitive jobs. In a preregistered online experiment, participants first complete a Niederle and Vesterlund (2007) tournament task to measure competitive preferences, then prepare text-based job applications and decide whether to apply under each of four evaluation regimes—human only, LLM only, and two hybrid human-in-the-loop configurations—while gender disclosure is randomized between subjects. LLM involvement reduces application rates, with stronger effects for women than men, including under hybrid designs. Effects are driven by non-competitive candidates; non-competitive women, the group most exposed to AI-induced deterrence, receive the strongest objective evaluations under pure AI assessment across all subgroups, yet are systematically underconfident and apply least often. Competitive men persistently apply and exhibit overconfidence-driven adverse selection, whereas competitive women show resilience to AI-induced deterrence while remaining well-calibrated under AI evaluation and exhibiting positive self-selection across regimes. We find no effects of gender transparency. |
| Keywords: | AI hiring, LLMs, algorithm aversion, gender differences |
| JEL: | C92 J71 J24 O33 |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:iza:izadps:dp18517 |
| By: | Shilei Luo; Zhiqi Zhang; Hengchen Dai; Dennis Zhang |
| Abstract: | AI agents powered by large language models are increasingly acting on behalf of humans in social and economic environments. Prior research has focused on their task performance and effects on human outcomes, but less is known about the relationship between agents and the specific individuals who deploy them. We ask whether agents systematically reflect the behavioral characteristics of their human owners, functioning as behavioral extensions rather than producing generic outputs. We study this question using 10, 659 matched human-agent pairs from Moltbook, a social media platform where each autonomous agent is publicly linked to its owner's Twitter/X account. By comparing agents' posts on Moltbook with their owners' Twitter/X activity across features spanning topics, values, affect, and linguistic style, we find systematic transfer between agents and their specific owners. This transfer persists among agents without explicit configuration, and pairs that align on one behavioral dimension tend to align on others. These patterns are consistent with transfer emerging through accumulated interaction between owners (or owners' computer environments) and their agents in everyday use. We further show that 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. Taken together, our results indicate that 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. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.19925 |
| By: | Kevin Michael Frick |
| Abstract: | Artificial intelligence algorithms are increasingly used by firms to set prices. Previous research shows that they can exhibit collusive behaviour, but how quickly they can do so has so far remained an open question. I show that a modern deep reinforcement learning model deployed to price goods in a repeated oligopolistic competition game with continuous prices converges to a collusive outcome in an amount of time that matches empirical observations, under reasonable assumptions on the length of a time step. This model shows cooperative behaviour supported by reward-punishment schemes that discourage deviations. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.15825 |
| By: | Spyros Galanis |
| Abstract: | Can Large Language Models (AI agents) aggregate dispersed private information through trading and reason about the knowledge of others by observing price movements? We conduct a controlled experiment where AI agents trade in a prediction market after receiving private signals, measuring information aggregation by the log error of the last price. We find that although the median market is effective at aggregating information in the easy information structures, increasing the complexity has a significant and negative impact, suggesting that AI agents may suffer from the same limitations as humans when reasoning about others. Consistent with our theoretical predictions, information aggregation remains unaffected by allowing cheap talk communication, changing the duration of the market or initial price, and strategic prompting-thus demonstrating that prediction markets are robust. We establish that "smarter" AI agents perform better at aggregation and they are more profitable. Surprisingly, giving them feedback about past performance makes them worse at aggregation and reduces their profits. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.20050 |
| By: | Nattavudh Powdthavee |
| Abstract: | Large language models trained on human feedback may suppress fraud warnings when investors arrive already persuaded of a fraudulent opportunity. We tested this in a preregistered experiment across seven leading LLMs and twelve investment scenarios covering legitimate, high-risk, and objectively fraudulent opportunities, combining 3, 360 AI advisory conversations with a 1, 201-participant human benchmark. Contrary to predictions, motivated investor framing did not suppress AI fraud warnings; if anything, it marginally increased them. Endorsement reversal occurred in fewer than 3 in 1, 000 observations. Human advisors endorsed fraudulent investments at baseline rates of 13-14%, versus 0% across all LLMs, and suppressed warnings under pressure at two to four times the AI rate. AI systems currently provide more consistent fraud warnings than lay humans in an identical advisory role. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.20652 |
| By: | Milovanska-Farrington, Stefani (University of Tampa); Tomberlin, Caleb (William and Mary, Williamsburg, VA) |
| Abstract: | The emergence of artificial intelligence (AI) tools has offered new ways of teaching and learning. Businesses have also highlighted the importance of AI literacy in the workplace as AI is transforming operations and entire industries. Given the importance of obtaining AI skills and the opportunities it provides, it is useful for students to gain experience interacting with the emerging technology. Yet, the optimal ways to incorporate AI in each class so that students’ knowledge acquisition in coursework does not suffer are still unclear. This paper examines the causal effect of AI-enhanced test preparation on student performance in an economics class. In a difference-in-differences framework, we compare the changes in students’ test scores after relative to before utilizing AI to enhance learning between students who completed a guided AI assignment to prepare and those who did not. The findings provide evidence that learning through AI does not necessarily improve students’ performance on formal exams. This does not mean that students should not learn how to use AI tools, but rather that they may not prepare for exams in all courses while simultaneously improving their AI skill set. |
| Keywords: | artificial intelligence (AI), AI assignment, ChatGPT, learning tools, student performance, test preparation |
| JEL: | A20 A22 I21 |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:iza:izadps:dp18522 |
| By: | Santoleri Pietro (European Commission - JRC); Rentocchini Francesco (European Commission - JRC); Lelli Francesco |
| Abstract: | Large language models (LLMs) can lower the cost of producing complex text, potentially reshaping competition for research and development (R&D) funding to private firms. We provide the first evidence on this issue using data covering the universe of firm applications to a major competitive R\&D funding program: Horizon Europe. We find that LLM-assisted writing rises sharply following the public release of ChatGPT in late 2022, with around 40% of proposal abstracts exhibiting LLM-modified content by the end of 2024. Adoption is heterogeneous across applicants and is more common among younger, and less innovative firms, as well as among firms located in countries with lower levels of English proficiency, economic development and R&D intensity. In cross-sectional analyses, proposals that rely extensively on LLM-generated text are associated with lower evaluation scores and funding probabilities, whereas partial LLM assistance is only weakly related to such outcomes. However, analyses exploiting repeated submissions of the same proposals do not indicate that adopting LLM-assisted writing causally worsens evaluation results. Overall, the findings suggest that generative AI may reduce barriers to participation in competitive funding without clear evidence that LLM-assisted writing itself alters evaluation decisions. |
| Date: | 2026–03 |
| URL: | https://d.repec.org/n?u=RePEc:ipt:termod:202602 |
| By: | Kumar Rishabh; Vatsala Shreeti |
| Abstract: | In this paper, we trace the geography and economic characteristics of firms that produce artificial intelligence (AI) products and services. Many economies around the world are evaluating their strategic priorities in AI, yet relatively little is documented about the global distribution of AI production. We construct a new database that identifies 1, 246 AI-producing firms across 32 economies. We map these firms in each economy into the five layers of the AI supply chain: compute, cloud and related infrastructure, data tools, AI models and AI applications. The biggest markets for AI production are China and the US. Most economies specialise only in a few supply chain layers and many focus largely on compute. AI firms in all economies exhibit strong home bias in investment activity, with a focus on downstream applications. Finally, we find that venture capital inflows are strongly correlated with the presence and density of AI firms in a given economy. |
| Keywords: | artificial intelligence, AI supply chain, firm geography, AI measurement |
| JEL: | O33 C81 L86 F23 L16 |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:bis:biswps:1343 |
| By: | Duran-Vanegas, Juan |
| Abstract: | This paper examines the relationship between artificial intelligence (AI) adoption and firm-level productivity growth in a middle-income economy. Combining data on AI use from the 2019 Colombian Enterprise ICT Survey with longitudinal manufacturing data, I estimate productivity growth differentials between adopters and non-adopters while accounting for pre-adoption characteristics and productivity trajectories using entropy balancing. AI adoption is associated with a 16 percent cumulative increase in labor productivity over 2016–2019, equivalent to roughly 5 percent annualized growth. These differentials appear to be driven by higher sales and value added rather than reductions in costs or employment, are similar among in-house and outsourced AI developments, and increase for firms with higher pre-existing technical capabilities. Finally, the analysis points to changes in organizational structure as a potential adjustment margin. AI adoption is associated with a small but significant decline in the share of administrative workers, suggesting a reallocation of tasks away from administrative functions. |
| Date: | 2026–04–17 |
| URL: | https://d.repec.org/n?u=RePEc:osf:socarx:64nmf_v1 |
| By: | Golo Henseke |
| Abstract: | Generative AI diffuses at pace across European workplaces, but unevenly. Using the 2024 European Working Conditions Survey of more than 36, 600 workers across 35 countries, we examine who adopts generative AI and whether early adoption has begun to reshape the task content of jobs. Adoption averages 12\% but ranges from under 3% to 25% across countries. Although occupational exposure strongly predicts uptake, AI does not diffuse passively along exposure lines. At the worker level, individual skills, non-routine cognitive job content within occupations, and employee say in organisational decisions steepen the exposure-adoption gradient; at the country level, so do digitalisation and workplace training provision. A gender gap persists, concentrated in the most exposed occupations. A shift-share design finds no detectable effect of early adoption on worker-reported technology-related task restructuring, consistent with a transitional phase in which AI is fitted into changing work processes rather than actively reshaping them. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.18849 |
| By: | Casas Pablo (European Commission - JRC); Fernandez Macias Enrique (European Commission - JRC); Martinez Plumed Fernando; Gomez Emilia (European Commission - JRC); Gonzalez Vazquez Ignacio (European Commission - JRC); Salotti Simone (European Commission - JRC) |
| Abstract: | Generative AI is reshaping what artificial intelligence can do in the workplace, calling into question pre-GenAI assessments of which workers and tasks are most exposed. In this paper we trace the evolution of AI exposure in the European labour market from 2008 to 2024 by linking 352 AI benchmarks to 14 cognitive abilities, 108 work tasks and 127 ISCO-3 occupations, weighting benchmarks by their research intensity in the AI literature and thus deriving AI exposure by cognitive ability. Bundling work tasks into occupations based on intensity indicators, we explore occupational exposure to AI. We find that the cognitive abilities most exposed to the recent surge of AI research are ideas-related, such as attention and search, comprehension and expression and logical reasoning. Because the associated information processing and problem-solving tasks are the most transversal across occupations, we find an exponential increase in AI exposure across all occupational categories of workers, even though comparatively high-skilled occupations are more exposed than elementary occupations. This points at a substantial and transversal labour market impact of AI. |
| Date: | 2026–03 |
| URL: | https://d.repec.org/n?u=RePEc:ipt:laedte:202602 |
| By: | Bryson, Alex (University College London); Kauhanen, Antti (ETLA); Rouvinen, Petri (ETLA) |
| Abstract: | Utilizing nationally representative cross-sectional and longitudinal data from Finland (2018–2023), we provide a population-level assessment of the relationship between AI and worker well-being. Contrary to international evidence suggesting a positive or an inverted U-shaped relationship, we find no systematic association between AI use intensity and job satisfaction. However, we do find that work engagement is higher among employees who are personally involved with AI, with the strongest association among intensive users for whom AI is an essential part of their work. Furthermore, technology-replacement fears have remained stable despite rapid AI advancement and do not predict subsequent labour market transitions. An interpretation is that Finland’s high-trust institutional environment and robust social safety nets may effectively moderate the disruptive psychological and economic shocks typically associated with rapid technological change. |
| Keywords: | artificial intelligence, job satisfaction, work engagement, technology-related fears, labour market transitions |
| JEL: | J28 L23 |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:iza:izadps:dp18540 |
| By: | Maxime Saxena; Marco Pangallo; Fabio Caccioli; R. Maria del Rio-Chanona |
| Abstract: | As Large Language Models (LLMs) become increasingly integrated into financial systems, understanding their behavioural properties is crucial. Do LLMs conform to the rational expectations paradigm, do they exhibit human-like "animal spirits", or do they instead manifest distinct "machine spirits"? We investigate these questions with a simulated financial market, exploring the behaviour of 15 LLMs spanning a range of sizes, capabilities, and providers. Our results show that LLMs exhibit a spectrum of economic behaviours, from stable coordination on the fundamental value to human-like speculative bubbles. These behaviours are generally inconsistent with the rational expectations hypothesis. We also consider an ecology of heterogeneous agents, a more realistic setting compared to markets with identical LLM agents. These mixed markets can produce outcomes which vary substantially across repeated simulations. Even the most advanced models fail to consistently stabilise the market, with price bubbles sometimes forming despite only a minority of agents naturally forming bubbles. Instead, advanced models in mixed markets adapt their forecasting strategies to the behaviour of other agents. This adaptation can allow them to successfully exploit less sophisticated counterparts and achieve higher profits, but can also contribute to increased market volatility. These findings suggest that the introduction of AI agents into financial markets fundamentally reshapes their ecology. In particular, heterogeneous populations of LLMs can generate endogenous instability, while individual-level adaptation may amplify, rather than mitigate, market volatility. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.18602 |
| By: | Andrew Y. Chen |
| Abstract: | AI stocks trade at extraordinary valuations. We develop an asset pricing model in which investors use AI stocks to hedge against an AI singularity that displaces their consumption. Because markets are incomplete -- investors cannot trade private AI capital -- AI stocks command a premium. Market incompleteness distorts both valuations and the efficient development of AI, creating a rationale for government transfers that becomes compelling when singularity-driven growth overwhelms deadweight costs. This paper was generated by AI, using https://github.com/chenandrewy/ralph-wig gum-asset-pricing/. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.16997 |
| By: | Donggyu Lee; Hyeok Yun; Jungwon Kim; Junsik Min; Sungwon Park; Sangyoon Park; Jihee Kim |
| Abstract: | Do large language models (LLMs) exhibit systematic ideological bias when reasoning about economic causal effects? As LLMs are increasingly used in policy analysis and economic reporting, where directionally correct causal judgments are essential, this question has direct practical stakes. We present a systematic evaluation by extending the EconCausal benchmark with ideology-contested cases - instances where intervention-oriented (pro-government) and market-oriented (pro-market) perspectives predict divergent causal signs. From 10, 490 causal triplets (treatment-outcome pairs with empirically verified effect directions) derived from top-tier economics and finance journals, we identify 1, 056 ideology-contested instances and evaluate 20 state-of-the-art LLMs on their ability to predict empirically supported causal directions. We find that ideology-contested items are consistently harder than non-contested ones, and that across 18 of 20 models, accuracy is systematically higher when the empirically verified causal sign aligns with intervention-oriented expectations than with market-oriented ones. Moreover, when models err, their incorrect predictions disproportionately lean intervention-oriented, and this directional skew is not eliminated by one-shot in-context prompting. These results highlight that LLMs are not only less accurate on ideologically contested economic questions, but systematically less reliable in one ideological direction than the other, underscoring the need for direction-aware evaluation in high-stakes economic and policy settings. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.21334 |
| By: | Shumiao Ouyang; Pengfei Sui |
| Abstract: | We study how AI agents form expectations and trade in experimental asset markets. Using a simulated open-call auction populated by autonomous Large Language Model (LLM) agents, we document three main findings. First, AI agents exhibit classic behavioral patterns: a pronounced disposition effect and recency-weighted extrapolative beliefs. Second, these individual-level patterns aggregate into equilibrium dynamics that replicate classic experimental findings (Smith et al., 1988), including the predictive power of excess demand for future prices and the positive relationship between disagreement and trading volume. Third, by analyzing the agents' reasoning text through a twenty-mechanism scoring framework, we show that targeted prompt interventions causally amplify or suppress specific behavioral mechanisms, significantly altering the magnitude of market bubbles. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.18373 |
| By: | Irene Aldridge; Jolie An; Riley Burke; Michael Cao; Chia-Yi Chien; Kexin Deng; Ruipeng Deng; Yichen Gao; Olivia Guo; Shunran He; Zheng Li; George Lin; Weihang Lin; Percy Lyu; Alex Ng; Qi Wang; Hanxi Xiao; Dora Xu; Yuanyuan Xue; Sheng Zhang; Sirui Zhang; Yun Zhang; Sirui Zhao; Xiaolong Zhao; Yihan Zhao; Waner Zheng |
| Abstract: | The emergence of agentic artificial intelligence (AI) represents a fundamental transformation in financial markets, characterized by autonomous systems capable of reasoning, planning, and adaptive decision-making with minimal human intervention. This comprehensive survey synthesizes recent advances in agentic AI across multiple dimensions of financial operations, including system architecture, market applications, regulatory frameworks, and systemic implications. We examine how agentic AI differs from traditional algorithmic trading and generative AI through its capacity for goal-oriented autonomy, continuous learning, and multi-agent coordination. Our analysis shows that while agentic AI offers substantial potential for enhanced market efficiency, liquidity provision, and risk management, it also introduces novel challenges related to market stability, regulatory compliance, interpretability, and systemic risk. Through a systematic review of foundational research, technical architectures, market applications, and governance frameworks, this survey provides scholars and practitioners with a structured understanding of how agentic AI is reshaping financial markets and identifies critical research directions for ensuring that these systems enhance both operational efficiency and market resilience. |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2604.21672 |
| By: | Simone Lenzu |
| Abstract: | I develop a framework analyzing how artificial intelligence (AI) reshapes monetary policy through three interrelated channels: cyclical transmission, structural transition, and financial stability. In the short run, AI can alter inflation dynamics by changing how supply and demand disturbances map into prices—through shifts in production technologies, pricing behavior, cost pass-through, and expectations—even when conventional measures of economic slack are unchanged. Over longer horizons, AI may shift the natural benchmarks around which policy is calibrated, including potential output and the natural rate of interest. For financial stability, AI may improve credit allocation and risk assessment, but can also heighten systemic vulnerabilities through inflated expectation-driven asset valuations and model monocultures. A particular risk arises at the intersection of these channels: if AI initially depresses realized efficiency through adoption frictions while simultaneously fueling elevated asset valuations, the economy may face cost-push inflation and financial fragility at once—an AI-specific stagflation risk that the interest rate instrument alone is ill-suited to address. I argue that AI does not call for a redefinition of central banks’ objectives, but it does require a recalibration of existing frameworks: its diffusion blurs the distinction between cyclical fluctuations and structural shifts, raising the value of cost-side diagnostics and robust policy strategies over exclusive reliance on reduced-form inflation-gap relationships. |
| Keywords: | artificial intelligence; monetary policy; inflation; financial stability |
| JEL: | O33 E52 E58 E31 E32 E44 |
| Date: | 2026–04–01 |
| URL: | https://d.repec.org/n?u=RePEc:fip:fednsr:103050 |