Schedule
November 4, 2025 / location: ST Center, Seoul, Korea
This workshop is planned as a one-day event featuring invited talks from leading researchers, peer-reviewed paper presentations, a specialized session for Korea R&D result sharing session and a panel discussion. See the workshop program schedule below.
Last Updated on Oct. 24
We are excited to announce that our program is fully integrated into the main conference’s technical program! All 9 selected AID Workshop papers will be presented in a dedicated technical session open to all conference attendees.
This year’s program features 9 high-quality, peer-reviewed papers divided into two specialized sessions:
- Session I: System Dependability. This session focuses on the robustness, verification, and simulation of dependable systems, including research on cyber-deception, robustness evaluation, and attack path modeling.
- Session II: LLM-Driven Dependability. This session dives into the role of Large Language Models in security, featuring work on automated vulnerability repair, forensic analysis, and autonomous security testing frameworks.
| Time (KST) | Event |
|---|---|
| 09:00–09:10 | Opening remark for AID’25 |
| 09:10–10:50 | AID Workshop Session I—System Dependability |
| 11:10–12:30 | AID Workshop Session II—LLM-Driven Dependability |
| 12:30–14:00 | Lunch |
| 14:00–15:00 | PRDC session |
| 15:10–16:00 | AID Workshop Keynote— “How to write secure programs with LLMs?” Prof. Hyoungshick Kim (Sungkyunkwan University) |
| 16:30–17:30 | PRDC session |
| 18:30–20:00 | Banquet |
Presentation schedule
Each presentation is allotted a 20-minute slot, structured as a 15-minute talk followed by a 5-minute Q&A session.
Session I
09:10–09:30 Robustness Evaluation under RGB-Camera Attacks in CARLA: A Systematic Evaluation of Color Modes and Attack Types)
09:30–09:50 A Study on User Approval Criteria and User Interfaces for Dependability of AI-based Code Generation Tools
09:50–10:10 A Self-Synchronizing Cyber Deception Framework via Infrastructure as Code Reflection
10:10–10:30 An approach to creating the optimal attack path based on reinforcement learning
10:30–10:50 Toward Dependability Simulation for Intelligent Transportation Systems in Connected Vehicle Environments
Session II
11:10–11:30 Automated Vulnerability Repair of Obfuscated and Non-Obfuscated Smart Contracts Using Large Language Models
11:30–11:50 Automated Timeline-Based Forensic Report Generation with Anomaly Detection and LLM-Based CTI Mapping
11:50–12:10 Dependable Code Repair with LLMs: AI-Driven Vulnerability Detection and Automated Patching
12:10–12:30 Toward an Autonomous Purple Teaming Framework for Security and Safety in Large Language Models
Keynote
15:10–16:00 How to write secure programs with LLMs?