Knowledge that Transforms

While conventional software is programmed for specific tasks, AI learns and refines its capabilities through a two-stage process: pre-training and fine-tuning. A central feature of this process is the AI adaptive learning embedded in the fine-tuning stage, whereby user interactions refine the AI’s capabilities initially developed from publicly available data during pre-training. This feature reshapes the firm’s freemium decision. We develop an analytical model to examine how AI adaptive learning affects firms’ optimal freemium strategies and find that while a free version can expand the user base and generate learning data, it also intensifies cannibalization of demand and increases the service-cost burden. We demonstrate that highly effective adaptive learning can paradoxically harm profitability by excessively enhancing the free version’s appeal and sharply cannibalizing premium demand. We find further that firms may optimally avoid offering a free version even when service costs are minimal, relying instead on reductions to the price of premium service to expand the paid user base. Narrowing the base capability gap between versions can also increase profitability by strengthening adaptive learning despite the increased cannibalization. Finally, results show that introducing a free version does not necessarily increase consumer surplus, because improved AI capabilities may allow firms to raise premium prices. These results are robust across several model extensions and inform both firms’ monetization strategies and policy discussions of consumer welfare in generative AI markets.

Designing effective security service contracts presents a critical challenge for both Managed Security Service Providers (MSSPs) and their clients, driven by heterogeneous client risks, interdependent security externalities, and post-adoption client behaviors. Employing a game-theoretic framework, this research investigates an MSSP’s optimal contract strategy among three regimes: a fixed compensation contract (F-Contract), a proportional compensation contract (R-Contract), and a menu-based contract regime (M-Contract) that offers both options simultaneously, allowing clients to self-select. The differences in compensation structures between the F-Contract and R-Contract directly shape client behavior, resulting in varying levels of post-adoption client negligence. The menu-based contract design in our setting introduces a richer mechanism than simply segmenting heterogeneous clients: clients’ strategic migration across contract options changes the aggregate security of the protected network and, consequently, the MSSP’s optimal market coverage and profitability. As security loss risk increases, high-valuation clients migrate from fixed to proportional compensation. When this migration reduces aggregate negligence, it offsets the intensified negative externalities and induces the MSSP to expand, rather than contract, its client base and profits. Crucially, this reversal cannot arise under either pure contract regime or in screening models without contract-dependent behavior. We further show that no single regime always maximizes market coverage, MSSP profitability, or social welfare, as the optimal choice depends on the security loss risk and the negligence gap between the F- and R-Contracts. Specifically, when both factors are moderate, the M-Contract yields a “win-win-win” equilibrium that simultaneously advances the interests of the MSSP, clients, and social planners, requiring no external policy intervention. However, we also find that an MSSP’s privately optimal contract does not always inherently maximize social welfare. To address this, we characterize the conditions that call for targeted regulatory or contractual interventions to realign private and social incentives.

For a collaboration between humans and artificial intelligence (AI) to be fruitful, tasks should be allocated based on their complementary capabilities. Prior research shows that when humans are responsible for allocating tasks between themselves and an AI through delegation, they often delegate too infrequently or delegate the wrong tasks, preventing complementary performance gains.We study how different types of AI system information affect both delegation frequency and delegation effectiveness, which capture the extent to which humans can leverage existing complementarities with AI. Specifically, we study ex-ante AI certainty (the AI’s estimated likelihood of being correct) and ex-post AI outcome information (whether the AI was actually correct on a given task). We show experimentally that presenting either AI certainty before or the AI’s outcome after a delegation decision has no or even negative effects on combined human-AI performance. However, providing both AI certainty and AI outcome information leads to increased delegation frequency as well as more effective delegation, ultimately leading to beneficial performance. We find that ex-ante certainty information calibrates users’ expectations about AI performance on the task-instance level, while ex-post outcome information confirms or disconfirms these expectations. This complementary use of AI system information supports more accurate mental models of the AI’s capabilities, reduces unwarranted algorithm aversion and improves appropriate task allocation. Overall, our results show that the effects of AI system information should not be assessed in isolation. While each signal on its own can be uninformative or even harmful, combining them can reverse the potentially harmful individual effects and facilitate effective human-AI collaboration. Our findings have implications for the design of AI systems in collaborative delegation settings, suggesting that carefully designed system information can help users better leverage complementarities with AI.

Recent regulatory scrutiny has highlighted concerns regarding Amazon’s use of market data in competing with its sellers through its private-label operations, prompting the introduction of new data governance policies across multiple jurisdictions. Motivated by this development, this paper examines the effectiveness of regulations governing platform data use. We study two policies: Policy S, which prohibits the retailer’s use of market data while granting exclusive access to sellers, and Policy RS, which allows shared data access between the retailer and sellers, reflecting emerging industry practices. These policies are evaluated against a baseline case with no data restrictions. Our analysis shows that while Policy S effectively restricts platform data use and protects sellers, it also generates strong incentives for the retailer to engage in strategic responses, including contractual adjustments and policy circumvention, as it represents the least favorable outcome for the platform. Policy RS, in contrast, emerges as a practical compromise that better aligns platform and seller incentives. We further show that the effectiveness of data regulations depends on the contractual environment and the platform’s ability to adjust its organizational structure. A key insight is that the retailer can undermine regulatory intent through subtle but systematic strategies, such as steering sellers across contract forms and exploiting regulatory ambiguities. These effects become more pronounced under the wholesale contract. Our findings highlight that recent regulations may be insufficient if they do not account for strategic platform behavior. We find that recent data regulations face important challenges, as they may not prevent the retailer from exploiting contractual flexibility and regulatory loopholes. We therefore provide guidance for designing more robust regulatory frameworks with clearer conditions and enforcement mechanisms that explicitly incorporate contract structure and limit opportunities for regulatory circumvention, thereby strengthening protection against unfair competitive practices by dominant platforms.

Rating inflation is prevalent in reputation systems and can reduce their informational value. Through an exploratory mixed‐method study, we examine how and why rating inflation behavior evolves over time as reviewers socialize within an online reputation system platform community. We show that some reviewers are more prone to inflation while others avoid it. We draw on the theory of reciprocity and distinguish among three archetypical phases of reviewers: (1) the Newcomer Phase, (2) the Inflator Phase, and (3) the Veteran Phase. Newcomers who are not yet socialized in the platform community are less likely to inflate their ratings. Inflators exhibit the social norm of direct reviewer reciprocity and inflate their ratings. Veterans, reflecting the social norm of generalized reviewer reciprocity, are less likely to inflate ratings. Instead, they are more candid because they are driven by a commitment to help the platform community. These archetypical phases correspond to reviewers’ increasing socialization within the platform community and exposure to different digital platform affordances. Together, these processes shape which forms of reciprocity become salient over time—whether lacking, direct, or generalized—which consequently drive distinct reviewer behaviors. While most research on rating inflation emphasizes direct reciprocity, we highlight how reviewers can shift toward less inflated behaviors over time, aligning with generalized reciprocity towards the platform community. Thus, this exploratory work suggests an evolutionary process of rating behavior whereby rating inflation follows an inverted U-shape over the reviewer lifecycle.

This research examines how externally hired senior technology executives reshape product innovation in startup complementors operating within software platform ecosystems. We distinguish two innovation types tied to the platform’s layered modular architecture: layer-expansion innovation that extends products into previously unused technical layers, and within-layer innovation that deepens the use of layers already embedded in the complementor’s products. Using longitudinal data from startup complementors in the Hadoop ecosystem, we find that externally hired senior technology executives are positively associated with layer-expansion innovation but negatively associated with within-layer innovation, indicating a reallocation of innovation efforts across platform layers. Drawing on the cognitive and technical dimensions of tacit knowledge, we theorize that externally hired senior technology executives influence innovation by introducing knowledge from their prior experience in technology and product development. Multiple mechanism analyses provide evidence consistent with this knowledge-based explanation. The innovation effects are stronger when executives are hired into newly created positions and when their prior expertise extends beyond the startup’s existing technical layers. We also find that these effects are more pronounced among growth-stage startups and startups without senior technical founders. This research contributes to the software platform ecosystem, executive human capital, and entrepreneurial innovation literature by investigating the impact of externally hired senior technology executives on startup innovation relative to the platform’s layered architecture.

eHealth behavioral interventions have transformed how individuals manage their health and modify their lifestyles. Despite their growing popularity, many users gradually reduce or discontinue their participation over time. To understand this disengagement, we extend the Expectation-Confirmation Theory (ECT) by modeling user engagement as a dynamic learning process, where evolving perceptions of intervention effectiveness are shaped by ongoing experience. Leveraging a hierarchical Bayesian learning framework, we analyze users’ perception updates and how such dynamics influences engagement decisions. Our empirical results show that individuals’ learning performance appears to be lower for interventions with ambiguous instructions or those focused on short-term health outcomes. These interventions tend to generate noisier feedback that may hinder users’ ability to form accurate perceptions of intervention effectiveness, which may ultimately reduce sustained engagement. Given that eHealth behavioral interventions typically possess credence characteristics, where effectiveness may be difficult for users to evaluate directly, the need for clear, accessible information becomes especially critical. To help improve user learning and engagement, we propose several denoising strategies and evaluate them through counterfactual simulations. Our work extends ECT into healthcare settings and provides actionable insights for designing more supportive Health IT systems that foster informed decision-making and sustain user engagement.

A decentralized autonomous organization (DAO) is a novel form of blockchain-based organization designed for collective decision-making. As DAOs emphasize a decentralized, democratic decision-making approach, participation serves as the foundation for their sustainable operation and development. Unfortunately, many DAOs struggle with low participation rates, often falling short of the required quorum. To address this critical issue, an increasing number of DAOs have adopted delegated voting, which allows members to transfer their voting rights to others. However, the impact of delegated voting within the DAO context remains unknown. By leveraging variation in the adoption of delegated voting across DAOs, we find that delegated voting increases members’ participation in proposal voting and enhances decision quality. Our results further show that delegated voting stimulates greater participation in proposals with higher participation costs, including those that are more complex, urgent, or operational in nature. However, in the long term, delegated voting also leads to greater voting power concentration and reduces engagement from both new and active voters, potentially harming sustained participation and the growth of the DAO community. Overall, our findings highlight the need for DAOs to balance the short-term gains from higher participation with the potential long-term risks to decentralization.

Algorithmic management (AM) has become a defining feature of online labor platforms (OLPs), profoundly shaping platform workers’ control over their working conditions. Prior research has documented diverse forms of worker resistance to AM—or algoactivism—yet existing studies rest on two problematic assumptions. First, that algoactivistic practices are uniformly accessible and arise directly from workers’ perceptions of structural constraints. Second, that such practices are primarily reactive resistance broadly targeted at the OLP’s AM system. These assumptions obscure heterogeneity in workers’ motivations and resources, as well as variation in how algoactivistic practices unfold. This study develops a more textured understanding of platform workers’ algoactivism by tracing how corresponding practices emerge through situated, reflective, and resource-dependent processes. Drawing on the contested terrain lens from labor process theory, we conceptualize the interplay between OLPs and workers as an ongoing struggle over control of working conditions. We examine this struggle in the context of Uber, a widely recognized extreme case of AM. Using a computer-assisted grounded theory approach that integrates topic modeling and qualitative coding procedures across multiple data sources, we develop a process-theoretical model of how platform workers contest AM. Our model centers on three recurring dynamics—reassessing terrain, exploring opportunities for contestation, and contesting terrain through algoactivistic practices—and yields two core theoretical contributions. First, we show that worker algoactivism depends on continual terrain reassessments and uneven capacities to engage in three forms of resourcing—algorithm, market, and voice resourcing. Second, we theorize algoactivism as a heterogeneous and multi-arena phenomenon comprising self-optimizing, distancing, and confronting practices that vary in logics, targets, and durability. Together, these contributions advance a more dynamic and agentic understanding of worker algoactivism and provide actionable insights for the design and governance of platform-mediated work.

Generative Artificial Intelligence (AI) technologies have emerged as a transformative force in the content creator economy. This paper examines how the AI stances signaled by specific platform policy decisions shape creator behavior on visual arts platforms. We leverage two independent natural experiments on leading Chinese platforms: Lofter’s launch of an AI image generator, which signaled a pro-AI stance, and Graffiti Kingdom’s prohibition of AI-generated artwork, which signaled an anti-AI stance. Our analysis shows that creators decreased their activity on Lofter following the AI generator launch, while activity increased on Graffiti Kingdom after the AI prohibition. Multiple lines of evidence show these effects stem from policy-communicated stance signaling: creators respond to what the focal AI policy communicates about AI’s role, the platform’s commitment to human creators, and the future competitive environment for human-created work, rather than merely to specific tool features or enforcement actions. Heterogeneity analysis reveals that higher-popularity, multi-homing, and AI-averse creators show larger activity reductions on Lofter. Through analysis of creator posts, we identify three primary concerns driving resistance: replacement risk, perceived low quality, and copyright infringement. To our knowledge, this is the first paper to causally identify how the AI stance signaled by a platform policy decision affects creator behavior. Our findings have implications for how platforms communicate AI-related policy decisions, and for policymakers on fostering a constructive relationship between AI and human creators.