About

Rohit Aggarwal is a Professor in the Department of Operations and Information Systems at the David Eccles School of Business, University of Utah. His research explores how AI technologies and human expertise can mutually enhance each other in organizational settings. His work is categorized under two main themes: AI augmenting human decision-making and humans augmenting AI decision-making. In the first theme, he examines AI's role in improving learning, skill development, and productivity within organizations, including field experiments on AI tools' impact on productivity and skill acquisition, as well as addressing global challenges such as linguistic divides through AI-powered coding tools. The second theme focuses on improving AI systems by embedding human insights to enhance performance and transparency, including developing explainable AI models for recruitment and hybrid frameworks for Generative AI that incorporate human guidance. Aggarwal's research aims to make AI systems more reliable, contextually aware, and aligned with real-world complexities, emphasizing the importance of human oversight and expertise in AI development and application. He holds a Ph.D. in Operations and Management Information Systems from the University of Connecticut and has received honors such as the Emerging Scholar Award from the University of Utah and a Doctoral Dissertation Fellowship. His scholarly contributions include numerous publications in top journals, and he has served in various academic and departmental roles, including workshop chair and department service.

Research topics

• Computer Science
• Political Science
• Psychology
• Business
• Marketing
• Knowledge management
• Finance

Selected publications

• Artificial Intelligence in Business Research: A Taxonomy of Methodological Roles and Future Directions

  SSRN Electronic Journal · 2026-01-01

  preprintOpen access
  PublisherDOI

• Self-Driving Technology and Its Acceptance: Explaining Differences in Reactions of the Heterogeneous Population to Potential Policies

  Production and Operations Management · 2025-04-04 · 3 citations

  article

  Self-driving vehicles have the potential to drastically reduce accidents caused by human errors, saving significant amounts of money in damages as well as human lives. However, public acceptance of the technology operating on public roads still needs to improve, as most Americans are uncomfortable sharing the road with a self-driving vehicle. The challenge for policymakers is to craft regulations that not only enhance the safety of self-driving technology but also foster public trust and acceptance. This study examines how specific policies—requiring visual cues to indicate when a vehicle is operating in self-driving mode and certification requirements for users—impact public acceptance of self-driving vehicles. To evaluate the impact of the policies, we theorize how policies may influence people's trust and how trust, in turn, may affect acceptance of the technology. Furthermore, we examine how these effects vary across political affiliations, as prior research suggests that Republicans and Democrats differ in their trust in government oversight and technological innovation. Our findings confirm that Republicans are generally less willing to share the road with self-driving vehicles than Democrats, largely due to lower trust in the government to regulate the technology effectively. We find that a visual cue policy increases trust in government but decreases trust in the technology, leading to increased acceptance among Republicans but a neutral or negative effect for Democrats. Conversely, a certification requirement increases trust in government and in other drivers, positively impacting acceptance for both Republicans and Democrats. Finally, additional analysis revealed that a combined policy implementing both measures proves to be the most effective at increasing overall public acceptance by strengthening trust across multiple dimensions. These insights provide valuable guidance for policymakers seeking to improve the integration of self-driving vehicles into public roadways.

  PublisherDOI

• Implementation challenges faced by artificial intelligence for national security

  2024-10-02

  book-chapter

  The essay “Computing Machinery and Intelligence,” authored by American computer scientist Alan Turing in the 1950s, introduces the concept of artificial intelligence (AI). “The ability of a computer system to do tasks that typically require human intelligence” is the definition of AI. The important fields of AI are automation, machine learning, and deep learning. AI began to rapidly develop in both the military and civilian sectors during the 1990s to increase national security. Work productivity and capacity are increased with the application of AI in India. Platforms or specific equipment can incorporate AI technology so that it can perform on its own. However, there are a few obstacles to overcome when implementing AI to provide security to the nation. Addressing implementation challenges requires a multidisciplinary approach involving collaboration between policymakers, domain experts, AI researchers, and ethicists. Implementing AI for national security must prioritize transparency, accountability, privacy protection, and human oversight to ensure responsible and effective deployment of AI technologies in safeguarding national interests.

  PublisherDOI

• Overcoming Limitations of Ai Agents: Integrating Tacit Knowledge Through Inferred Latent Themes

  SSRN Electronic Journal · 2024-01-01 · 1 citations

  articleOpen access1st authorCorresponding
  PublisherDOI

• The Effect of Gender Expectations and Physical Attractiveness on Discussion of Weakness in Online Professional Recommendations

  Information Systems Research · 2023-03-30 · 3 citations

  article1st authorCorresponding

  Companies are using online professional networks at an increasing rate to find qualified candidates to interview for job openings. Although recommendations published on these sites can provide valuable information and influence hiring decisions, the information may suffer from credibility issues due to the medium by which it is shared. In this study, we investigate whether including a discussion of a candidate’s weakness in a recommendation may be an effective way to increase the perceived credibility of the recommender and thereby improve the candidate’s chance of receiving an interview. We surveyed hiring managers and recruiters to collect data to measure the impact different recommendations have on their decisions. Our findings show that including a discussion of weakness in a recommendation increases the perceived credibility of the recommender, which has a positive effect on the candidate’s likelihood of being interviewed. However, when the discussion of weakness counters common gender-based expectations, it is harmful. When the discussion of weakness is consistent, it is helpful. Furthermore, we find that the physically attractive candidates (as shown in their profile picture) are harmed regardless of the weakness discussed. We investigate this further and find that additional discussion of the candidate’s strengths can reduce the negative impact of the discussion of weakness, but only if the strengths are consistent with common gender-based expectations.

  PublisherDOI

• Effect of Online Professional Network Recommendations on the Likelihood of an Interview: A Field Study

  Information Systems Research · 2023-03-30

  article1st authorCorresponding

  Online professional networks (OPNs) are an increasingly common tool used by recruiters to find and vet qualified job candidates for open positions. These sites allow users to publish recommendations given by other users to supplement their profile information and add credibility to the information provided. OPN recommendations offer a rich source of information to recruiters. Unlike recommendations shared in other ways (non-OPN recommendations), OPN recommendations are publicly accessible, and candidates have complete control over which recommendations they show to others. In this study, we investigate how recommendations may have a different effect when presented as an OPN recommendation versus a non-OPN recommendation. Furthermore, we explore how to improve the effectiveness of recommendations. We conducted a field study where we leveraged the candidate tracking system of a large recruitment firm to measure the varying impact recommendations have on recruiters’ decisions. Our findings show that OPN and non-OPN recommendations that discuss an expected weakness positively affect a candidate’s likelihood of being interviewed. In contrast, recommendations that discuss an unexpected weakness have a negative effect. Furthermore, we find that non-OPN recommendations with no discussion of weakness are significantly more effective than OPN recommendations with no discussion of weakness. We then show that the potential benefits of discussing an expected weakness are more pronounced for OPN recommendations, whereas the potential harm of discussing an unexpected weakness is more severe for non-OPN recommendations.

  PublisherDOI

• PRIVACY & SECURITY OF IOT

  2023-12-01

  book-chapter1st authorCorresponding

  The Internet of Things (IoT) is a network of devices that share information and coordinate their actions without human intervention. It is used in various industries, including manufacturing, healthcare, logistics, energy, and agriculture. IoT systems consist of wireless networks, cloud databases, sensors, data processing software, and networked smart devices. The functionality and implementation of IoT depend on the industry, but businesses must invest significant resources in cybersecurity. IoT authentication helps establish trust in IoT machines and devices, protecting data and managing access. Cost, convenience, and practicality are also important factors for IoT businesses. Data protection and asset management are crucial, with user privacy protection being more important than ever. Identity and access management (IAM) can help businesses set procedures to protect themselves from cyberattacks and data leaks. The IoT has become a highly influential sector in the tech industry, providing vital services like automatic home appliance management, healthcare monitoring, and transportation control. However, it also has security and privacy concerns. IoT data can be used for academic and research purposes, but data analysis must first remove identifying or private information to protect confidentiality. Users' privacy can be protected throughout the data processing lifecycle, with anonymous data collecting methods becoming increasingly common. Cloud storage services often store IoT data, and encryption methods, such as homomorphic encryption, are used for secret analysis and retrieval.

  PublisherDOI

• Differential Impact of Content in Online Communication on Heterogeneous Candidates: A Field Study in Technical Recruitment

  Information Systems Research · 2022 · 6 citations

  1st authorCorresponding
  • Computer Science
  • Knowledge management
  • Computer Science

  Recruitment is a critical activity for companies, and companies often communicate how they value their employees along with job requirements to potential candidates in a bid to attract them. However, there is an overall lack of understanding of how candidates react to such information and how their motivation toward the job changes with such online communication. Although there is substantial work that examines the decision-making process of managers who do technical hiring, to the best of our knowledge, there is a paucity of work that investigates the decision-making process of technical candidates. The broad research question studied is how including certain content in online communication about a technical job opportunity may (de)motivate heterogeneous candidates differently in applying for the job. We capture mediating variables, such as candidate prior performance and candidate experience level, that influence the effect of different online content on candidates’ propensity to apply and on candidates’ minimum acceptable salary increase. By testing actual job application behavior in a field study, we find that content related to employee work efforts or personal interests can attract high performers while discouraging low performers from applying in different contexts.

  PublisherDOI

• Superlatives and Scope of Improvement in Online Recommendations: Breath of Life or a Kiss of Death?

  MIS Quarterly · 2021 · 6 citations

  1st authorCorresponding
  • Computer Science
  • Computer Science
  • Psychology

  Online professional networks are important tools used by recruiters to find qualified candidates for job openings. Within these networks, professional recommendations are used to supplement profiles and add credibility. These recommendations tend to be overly positive, full of superlatives, and lacking in critical statements (referred to as scope of improvement). We draw on the theory of online trust to argue that having scope of improvement and superlatives may affect various dimensions of trust and to show how online trust, in turn, can affect the usefulness of a recommendation and the likelihood of receiving an interview. We contribute to the body of work on online trust both theoretically and empirically. From a theory perspective, we explain why including scope of improvement and superlatives in recommendations on online professional networks may help certain candidates in getting an interview but hurt others. From an empirical perspective, we provide a unique empirical setting that allows us to observe not only the effect of scope of improvement and superlatives, but also validate the theoretically argued underlying process. Furthermore, through discussion with recruiters, we identify then test contextual factors that differentiate recommendations on online professional networks from traditional recommendations. In this study, we use a scenario-based, quasi-experimental survey to test the effects of superlatives and scope of improvement on the usefulness and effectiveness of recommendations. Further, we test the mediating role of trust and how the experience levels of the recommendee affect the sign and strength of these relationships. Our findings indicate that including scope of improvement increases the effectiveness and usefulness of recommendations for candidates at low- and middle levels of experience. For the most experienced candidates, including scope of improvement has a negative effect on effectiveness. Superlatives negatively affect the perceived competence of the recommender and thus should be avoided. This negative effect is reduced when combined with scope of improvement.

  PublisherDOI

• Improving Funding Operations of Equity‐based Crowdfunding Platforms

  Production and Operations Management · 2021 · 21 citations

  1st authorCorresponding
  • Business
  • Finance

  Equity‐based crowdfunding platforms enable investors to come together to invest in startups and help lay‐investors to follow the lead of investors with good startup evaluation skills. Crowdfunding platforms often gather users’ inputs to evaluate investors and startups, but such inputs are quite noisy and often rely on past performance. Many investors with good evaluation skills do not have substantial past investment experience but still can lead investment rounds. This helps provide investment opportunities to lay‐investors who otherwise do not get to join investors with proven records. Without identifying such investors with potential, platforms lose the opportunity to put together investors to fund worthy startups and lose business. We develop a Bayesian model to address this problem and improve funding operations of equity‐based crowdfunding platforms. Specifically, the model helps platforms to better assess investors’ evaluation skills, identify lead investors for lay‐investors to follow, and increase funding opportunities on the platforms. To test the effectiveness of the proposed model, we gathered data from 319 actual investors listed on one of the largest crowdfunding platforms in the United States, picked startups randomly for investors to evaluate, and had investors evaluate startups in two ways—our approach and the conventional approach. We also discuss an extension of this Bayesian model that penalizes investors in case investors perform well by randomness. Furthermore, we used a Bayesian framework to help platforms better predict startup valuations accounting for investors’ evaluation skills.

  PublisherDOI

Frequent coauthors

• Harpreet Singh

  Wayne State University

  22 shared

• Irina Cojuharenco

  17 shared

• C. Emre Koksal

  The Ohio State University

  9 shared

• Philip Schniter

  9 shared

• Ram D. Gopal

  7 shared

• Vishal Midha

  Illinois State University

  5 shared

• Ara Darzi

  Johns Hopkins Medicine

  3 shared

• Alok Gupta

  3 shared

Labs

• Rohit Aggarwal LabPI

Education

• PhD, Operations and Information Management

  University of Connecticut

  2008

Awards & honors

• Emerging Scholar Award (2008)
• Doctoral Dissertation Fellowship Award (2008)