Biography

Dr. Rajiv Garg is an Associate Professor of Information Systems and Operations Management at Goizueta Business School at Emory University. His research examines how digital technologies, algorithms, and artificial intelligence shape human decision-making, learning, and economic outcomes. Using interdisciplinary methods from economics, statistics, and computer science, his work focuses on information flow in digital platforms, human–AI interaction, and the unintended consequences of algorithmic systems across domains such as education, marketing, labor markets, disaster response, and emerging technologies.

Dr. Garg’s research has been published in leading academic journals including Management Science, MIS Quarterly, Information Systems Research, Production and Operations Management, and the Journal of Management Information Systems, among others. His scholarship has been supported by more than $1 million in competitive research funding and has been presented widely at international conferences and universities across the globe. Beyond academia, his work has informed industry practice and public policy, and has received coverage in major media outlets such as Forbes, Fortune, CIO, and national news organizations.

At Emory, Dr. Garg plays a central leadership role in advancing AI education and research. He serves as Chair of the Goizueta AI Taskforce, Faculty Advisor to the Emory Center for AI in Learning (CAIL), and Faculty Advisor to the Goizueta Center for AI in Finance (GCAIF). He is also a member of the U.S. AI Safety Institute Consortium (US AISIC), where he contributes to national efforts on AI risk management, model evaluation, and safety standards. As ISOM PhD Program Area Coordinator, he oversees doctoral training, mentoring, and placement of PhD students.

Dr. Garg is an award-winning educator who has taught thousands of students across undergraduate, graduate, executive, and doctoral programs at Emory University, The University of Texas at Austin, and Carnegie Mellon University. His teaching has been recognized with multiple Distinguished Educator Awards, including Emory University’s Provost Distinguished Teaching Award for Excellence in Graduate and Professional Education. He is also a frequent keynote speaker and public intellectual, regularly engaging global audiences on artificial intelligence, analytics, and the future of work.

Media

  • ISOM 599R Data Management, Analysis, and Al
  • Meet Professor Rajiv Garg

Education

  • PhD in Information Systems & Management
    Carnegie Mellon University
  • M.Phil. in Public Policy & Management
    Carnegie Mellon University
  • MS in Computer Science (Databases & Networks)
    University of Southern California
  • M.S. in Electrical Engineering (Robotics & Intelligent Systems)
    University of Southern California
  • B.Tech. in Electrical Engineering
    Banaras Hindu University

Domain anchorage in LLMs: Lexicon profiling and unintended information leakage

Data & Policy
Lekha Challappa , Zijin Zhang, Rajiv Garg
October 27, 2025
This study investigates unintended information flow in large language models (LLMs) by proposing a computational linguistic framework for detecting and analyzing domain anchorage. Domain anchorage is a phenomenon potentially caused by in-context learning or latent “cache” retention of prior inputs, which enables language models to infer and reinforce shared latent concepts across interactions, leading to uniformity in responses that can persist across distinct users or prompts. Using GPT-4 as a case study, our framework systematically quantifies the lexical, syntactic, semantic, and positional similarities between inputs and outputs to detect these domain anchorage effects. We introduce a structured methodology to evaluate the associated risks and highlight the need for robust mitigation strategies. By leveraging domain-aware analysis, this work provides a scalable framework for monitoring information persistence in LLMs, which can inform enterprise guardrails to ensure response consistency, privacy, and safety in real-world deployments.