The 2nd SeT-LLM Workshop on Secure and Trustworthy Large Language Models
Workshop @ KDD 2026
Jeju, Korea | Workshop date TBA during August 9-13, 2026
Large language models are increasingly used in search, software engineering, scientific discovery, customer support, education, healthcare, and agentic systems. As these systems are deployed in more consequential settings, concerns about security, privacy, robustness, faithfulness, and safety become substantially more important. Failures such as prompt injection, jailbreaks, privacy leakage, hallucination, insecure tool use, and unreliable long-horizon behavior can create real downstream harm.
The SeT-LLM workshop focuses on the foundations and practice of building secure and trustworthy large language models. We aim to bring together researchers and practitioners from data mining, machine learning, natural language processing, systems, security, human-computer interaction, and applied domains to examine how LLMs can be evaluated, stress-tested, aligned, and deployed more responsibly.
The 2nd SeT-LLM Workshop will provide a venue for discussion on rigorous benchmarks, red teaming, privacy-preserving methods, trustworthy inference, robust agent design, and responsible applications of LLMs in high-stakes environments. Our goal is to advance technically grounded and practically useful approaches to dependable generative AI.
Workshop Goals
The workshop aims to foster a shared research agenda around secure and trustworthy LLMs. We focus on:
- Understanding failure modes in modern LLM systems, including security, safety, and privacy risks
- Developing rigorous evaluations for reliability, faithfulness, robustness, and trustworthiness
- Connecting deployment practice with research across alignment, governance, and high-stakes applications
