Sumit Kumar Kapoor

Dr. Sumit Kumar Kapoor

orcid
Researcher | Associate Professor & Dy. HOD in Computer Scien...

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 August 2025
Verified
Book Chapter
Closed access
NGNikhil Kumar Goyal
Nikhil Kumar Goyal
KGKapil Gulati
Kapil Gulati
AAAshish Avasthi
Ashish Avasthi
Sumit Kumar KapoorSumit Kumar Kapoor
Sumit Kumar Kapoor
BKBright Keswani
Bright Keswani
SGSangita Gupta
Sangita Gupta
Generative Artificial Intelligence (AI) has rapidly become a mainstream technology capable of autonomously producing realistic and diverse media content, including text, images, video, music, and code. As models like GPT, DALL·E, and Stable Diffusion transform industries such as entertainment, marketing, education, and software development, they also introduce a complex landscape of security and privacy risks. This chapter, per the authors, explores critical vulnerabilities such as prompt injection, adversarial attacks, data poisoning, and model inversion, which can lead to unethical outputs and exposure of sensitive training data. It further examines threats arising from large-scale data scraping, copyright violations, and lack of user consent, while also addressing ethical issues tied to misinformation, deepfakes, and surveillance. The chapter reviews technical safeguards including differential privacy, federated learning, watermarking, adversarial training, and secure deployment methods like sandboxing and real-time monitoring.

2 January 2026
Verified
Book Chapter
Closed access
PDF file
Sumit Kumar KapoorSumit Kumar Kapoor
Sumit Kumar Kapoor
KSKriti Sankhla
Kriti Sankhla
SKShikha Khullar
Shikha Khullar
UMUdit Mamodiya
Udit Mamodiya
VKVipin Khattri
Vipin Khattri

17 October 2025
Verified
Book Chapter
Closed access
PDF file
MSMadhav Sharma
Madhav Sharma
RDRitam Dutta
Ritam Dutta
DMDr. Richa Mathur
Dr. Richa Mathur
RJR. B. Hussana Johar
R. B. Hussana Johar
Sumit Kumar KapoorSumit Kumar Kapoor
Sumit Kumar Kapoor

14 November 2025
Verified
Book Chapter
Closed access
Sumit Kumar KapoorSumit Kumar Kapoor
Sumit Kumar Kapoor
KSKriti Sankhla
Kriti Sankhla
SSSaurabh Shandilya
Saurabh Shandilya
SASatya Prakash Awasthi
Satya Prakash Awasthi
Federated Learning (FL) allows for distributed model training across a population of devices while also offering the potential for privacy preservation, however, it remains susceptible to a range of adversarial manoeuvres, including model poisoning, backdoor attacks, and Byzantine behaviours. This chapter discusses secure multiparty computation (SMPC) as a cryptographic defence mechanism that allows models to be trained collaboratively and without revealing individual data associated with these models. We categorise the impacts of attacks, we offer various architectures of SMPC-based FL and we compare SMPC-based FL with traditional FL bases on security, communication, and computation. Certain key computational techniques, practical implementations with healthcare and industrial, Internet of Things (IoT) applications, security-performance trade-offs, scalability, limitations of current work, and an ordered taxonomy of threats to FL give insight into the performance and deployment of SMPC as a means to reduce threats to model accuracy.

3 February 2026
Verified
Book Chapter
Closed access
DCDeepali Chaudhary
Deepali Chaudhary
DDDebmitra Das
Debmitra Das
RDRitam Dutta
Ritam Dutta
Sumit Kumar KapoorSumit Kumar Kapoor
Sumit Kumar Kapoor
SSShilpa Singh
Shilpa Singh
GPGanesh Narayan Prajapat
Ganesh Narayan Prajapat