Understanding Prompt Injection Attacks in LLMs

Large Language Models (LLMs) are becoming an integral part of modern applications, but with great capabilities come new security challenges. One of the most important threats developers and AI practitioners need to understand is Prompt Injection.

Agenda for the hands-on workshop

In this session, we'll explore how prompt injection attacks work, why they are dangerous, and what developers can do to build safer AI-powered systems. Here's some hands on cover for the day --

• What prompt injection attacks are and why they matter, different types of prompt injection techniques used against LLMs

• Real-world examples of prompt injection vulnerabilities and how attackers bypass system prompts and safety instructions

• Hands on adaptation to the LLM workflow or agentic pentesting via Kali Linux during the workshop

• Inteactibe Quiz at the End of the Workshop with Swags for the participants.

Speaker For The Day:

We have two speakers for the day:-

1. Sagnik Roy: GDG JISU Core Team Cyber Security

2. Sayuj Sur: OWASP JISU Core Member for Cyber Security and Agentic Pentesting

Whether you're an AI enthusiast, developer, cybersecurity learner, or someone curious about the security side of artificial intelligence, this session will provide practical insights into one of the most important challenges in modern AI systems.

📅 Date:- 18th June 2026

⌚ Time:- 10:30 AM IST