Recent publications and ongoing projects

Jussupow, E., Bauer, K.,  Heigl, R., Vogt, B., & Hinz, O. (2026). The Indirect Disclosure Effect: How Disclosing Generative AI Use Impacts Human Creative Collaboration with AI (Information Systems Research, Forthcoming).

Regulators increasingly mandate transparency regarding generative AI (GenAI) use in creative work, aiming to protect audiences from deception while preserving creators' self-expression. One way of achieving this transparency is through disclosure labels that directly inform audiences about GenAI use. Yet, prior research focused almost exclusively on how such labels affect audience evaluations and paid surprisingly little attention to whether mandatory disclosure affects creators, too. We refer to this as the indirect disclosure effect. Drawing on Goffman's account of impression management, we theorize that creators who anticipate disclosure fear that audiences will not recognize their human creative agency, threatening their validation as a creative self, which leads them to adjust their collaboration with GenAI. To investigate this mechanism, we employ two nested mixed-methods experiments in which participants collaborate with a text-to-image GenAI tool under different disclosure conditions. We empirically establish the indirect disclosure effect: when disclosure is anticipated, the majority of creators withdraw from the creative process, leaving image generation to GenAI. We provide evidence that this withdrawal is driven by creators’ fears that audiences will not recognize their creative agency. Hence, the produced artifacts predominantly reflect computational rather than human creativity, which is also recognized and evaluated by the audience regardless of the direct disclosure label. Overall, our study reveals a fundamental tension at the heart of transparency regulation: by disclosing GenAI use through a simple label, regulators may inadvertently diminish the very human creative agency they aim to protect, and they do so before audiences ever see the label.

Yang, C., Bauer, K., Li, X., & Hinz, O. (2026). My Advisor, Her AI and Me: Evidence from a Field Experiment on Human-AI Collaboration and Investment Decisions. (Management Science 72(1), pp.242-264.).

Amid ongoing policy and managerial debates on keeping humans in the loop of artificial intelligence (AI) decision-making processes, we investigate whether human involvement in AI-based service production benefits downstream consumers. Partnering with a large savings bank in Europe, we produced pure AI and human–AI collaborative investment advice, which we passed to the bank customers and investigated the degree of their advice taking in a field experiment. On the production side, contrary to concerns that humans might inefficiently override AI output, our findings show that having a human banker in the loop of AI-based financial advisory by giving her the final say over the advice provided does not compromise the quality of the advice. More importantly, on the consumption side, we find that the bank customers are more likely to align their final investment decisions with advice from the human–AI collaboration, compared with pure AI, especially when facing more risky investments. In our setting, this increased reliance on human–AI collaborative advice leads to higher material welfare for consumers. Additional analyses from the field experiment along with an online controlled experiment indicate that the persuasive efficacy of human–AI collaborative advice cannot be attributed to consumers’ belief in increased advice quality resulting from complementarities between human and AI capabilities. Instead, the consumption-side benefits of human involvement in the AI-based service largely stem from human involvement serving as a peripheral cue that enhances the affective appeal of the advice. Our findings indicate that regulations and guidelines should adopt a consumer-centric approach by fostering environments where human capabilities and AI systems can synergize effectively to benefit consumers while safeguarding consumer welfare. These nuanced insights are crucial for managers who face decisions about offering pure AI versus human–AI collaborative services and also for regulators advocating for having humans in the loop.

Bauer, K., Grunewald, A., Hett, F., Jagow, J., & Speicher, M. (2026). Treatment Targeting by Scaled Behavioral Measurement. (Submitted to Journal of Political Economy).

We study how behavioral economics and machine learning can be combined to construct effective treatment targeting rules. In a large field experiment at a major German online fashion retailer (500,000 visitors), a loss-framing of a discount message has essentially zero average effect on purchases and returns. We implement a nested incentivized behavioral measurement experiment (N=582) to elicit individual loss aversion, a moderator for loss framing proposed by economic theory, and train a prediction model to impute loss-aversion categories from commonly observed digital footprints at scale. Targeting treatment based on this machine learning scaled behavioral measurement yields statistically significant revenue gains. By contrast, targeting based on a trained causal forests based delivers smaller and less robust improvements. The results illustrate how scaling behavioral measurement through machine learning methods can improve both the performance of algorithmic treatment assignment.

v. Zahn, M., Liebich, L., Jussupow, E., Hinz, O., Bauer, K. (2026). Knowing (not) to know: XAI and human metacognition. (Information Systems Research, Forthcoming, Best Paper Award WISE 2025)

The use of eXplainable artificial intelligence (XAI) to render black-box AI more interpretable to humans is gaining practical relevance. Prior research has shown that XAI can influence how humans "think''. Yet little is known about whether XAI also affects how people "think about their thinking,'' i.e., their metacognitive processes. We address this gap by investigating whether XAI affects metacognitive processes that guide human confidence judgments about their ability to perform a task and, thereby, their decision whether to delegate the task to AI. We conducted two incentivized experiments in which domain experts repeatedly performed prediction tasks, with the option to delegate each task to an AI. We exogenously varied whether participants initially received explanations about the AI’s overall prediction logic. We find that AI explanations improve how well humans understand their own ability to perform the task (metacognitive accuracy). This improvement causally increases both the frequency and effectiveness of human-to-AI delegation. Additional analyses show that these effects primarily occur when explanations reveal to humans that AI’s prediction logic diverges from their own, leading to a systematic reduction of overconfidence. Our findings highlight metacognitive processes as a central, previously overlooked channel through which AI explanations can influence human-AI collaboration. We discuss practical implications of our results for organizations implementing XAI to comply with regulatory transparency requirements as, for example, outlined in the EU AI Act.

Zacharias, J., v. Schenk, A., Klockmann, V., Knickrehm, C., Hinz, O., & Bauer, K. (2026). Decentralized Feature Selection for Machine Learning. (Minor Revision, European Journal of Information Systems)

In the age of machine learning (ML), consumers' personal data is widely used for personalized product recommendations. To address privacy concerns, regulations increasingly grant consumers control over their data. One implementation are ``opt-out of information use'' features that allow consumers to specify which of their collected personal data ML-powered recommender systems can harness. However, we conjecture that such features may have an unintended side effect: withholding data could inadvertently reveal insights about consumers' latent characteristics, thereby enhancing targeting possibilities. Through a controlled pre-registered experiment, we evaluate both consumers' perceptions and technical consequences of such opt-out features in the context of a typical search problem. Our results show that these features increase consumers' sense of control over the system and alleviate privacy concerns for those who actually withhold information. Paradoxically, withholding information can simultaneously be harnessed to improve the ML model's predictive accuracy. From a policy perspective, we highlight the need for additional regulations on how organizations may use information withholding decisions, particularly when consumers' interests conflict with those of the recommender provider.

Bauer, K., Nofer, M., Heinrich, B., Drachsler, H., Hinz, O. (2026). The Dependency Dilemma: How Machine Learning Decision Aids can Undermine Skill Growth. (Business Information Systems and Engineering, Forthcoming).

Advances in Machine Learning (ML) have led organizations to increasingly implement predictive decision aids to enhance employees’ decision-making performance. While such systems improve organizational efficiency in many contexts, they may inadvertently impact the development of human decision-making skills. Drawing on cognitive theories, this study examines how the use of ML-based decision aids impacts skill development and performance, particularly in scenarios where access to the system is disrupted, such as during system discontinuance, or when the system exhibits bias or errors. Using a novel experimental design tailored to address organizational challenges and endogeneity concerns, we identify causal effects of ML reliance on skill development in decision making. Specifically, we demonstrate that reliance on ML predictions can hinder the development of critical decision-making skills, resulting in significant performance drops when the system becomes unavailable. Furthermore, we find that the extent of trust in the system's predictions strongly influences the severity of this skill deficit. These findings highlight the need for thoughtful integration of ML decision aids, emphasizing the importance of balancing reliance with skill retention to mitigate risks associated with temporary or permanent system disruptions.

Bauer, K., Hett, F., Chen, Y., Kosfeld, M. (2024). Group Identity and Belief Formation: A Microfoundation of Political Polarization (Cond. Accept at The Economic Journal)

To evaluate the impact of group identity on belief formation, we conducted online experiments before and after the 2020 US presidential election. We elicit participants' beliefs about future unemployment statistics and health system rankings and provide relevant news summaries. We find that people pay money to avoid information from political outgroups and attribute lower weight to this information when updating beliefs. An intervention which unlabels information sources decreases outgroup information avoidance by 50%, with a significant impact on only groupish participants. A debiasing intervention equalizing instrumental values of information sources reduces only universalists' WTP for a private signal. The interventions have little effect on information processing. The evidence is consistent with a source-utility channel contributing to polarization.

v. Zahn, M., Güler, A., Bauer, K., Hinz, O. (2026). Echoes of the Ping: Indirect Effects of Mobile Notifications on App Engagement. (Submitted to Management Science)

Businesses increasingly employ mobile notifications to increase app engagement. Firms typically evaluate notification effectiveness by measuring whether users click a notification and thereby visit the app, a pathway we call the \emph{direct} effect. This paper examines whether businesses systematically underestimate notification effectiveness because notifications increase app engagement even when users do not click them. Specifically, we propose an underexplored \emph{indirect} effect, by which notifications create a memory trace of promoted content. When that content later becomes relevant, users recall it and open the app on their own, rather than by clicking the earlier notification. Because this indirect effect is difficult to isolate, we combine a field experiment comprising one million users of a customer-loyalty app with brick-and-mortar purchase data and a quasi-experimental shock from a mobile operating-system update. We find that the indirect effect is approximately twice as large as the direct effect and persists into the following day, well after the direct effect has faded. Importantly, the indirect effect is stronger for users encountering situational cues, such as making a purchase at a retailer featured in a prior notification, and occurs even after notifications were cleared from the device. These findings are theoretically relevant because they show that notifications shape app engagement not only through immediate attention but also through what users later recall from memory. They also inform notification policies, as we show that policies based on causal machine learning accounting for indirect effects are more effective than policies based on direct effects only.

Knickrehm, C., Reichmann, D., Ewertz, J., Bauer, K. (2026). Deep Listening: Managerial Persuasion Through Vocal Delivery in Earnings Calls. (Submitted to Information Systems Research)

Earnings calls constitute an important information source for investors and financial analysts. Managers carefully craft their content and wording to persuade market participants of favorable firm prospects. Beyond the literal content conveyed, research in psychology suggests that how individuals communicate through vocal delivery functions as a distinct persuasion cue that shapes recipients’ attitudes and beliefs. Building on this research, we develop a deep learning model that captures how confident managers sound as the primary channel of persuasive vocal delivery. Applying our deep learning model to a large sample of earnings call audio recordings reveals non-trivial and previously undocumented variation in managerial vocal confidence. We find that vocal confidence in scripted manager presentations is associated with future earnings surprises, suggesting that managers deliberately adjust their vocal delivery to convey genuine information. Consistently, stock prices increase and analysts revise earnings forecasts upward following manager presentations conveyed with a more confident vocal demeanor. Shedding light on a key mechanism underlying vocal persuasion, we perform a micro-structure analysis of manager presentations and find that analysts exhibit greater cognitive engagement with individual statements that managers deliver more confidently. This study contributes to communication analytics by uncovering vocal confidence as a previously hidden yet economically meaningful layer of managerial persuasion.