GDG on Campus Universiti Teknologi Malaysia - Johor Bahru, Malaysia
As AI Agents become more powerful, security becomes non-negotiable. What happens when an agent is manipulated to bypass safety barriers? These are called jailbreak attacks — and they’re one of the biggest risks in today’s AI systems. In this hands-on 90-minute codelab, you won’t just learn about AI security — you’ll build it yourself.
0 RSVP'd
Building Secure Multi-Agent Systems — Safety Guardrails Against Jailbreaks with Google ADK
As AI Agents become more powerful, security becomes non-negotiable.
What happens when an agent is manipulated to bypass safety barriers?
These are called jailbreak attacks — and they’re one of the biggest risks in today’s AI systems.
In this hands-on 90-minute codelab, you won’t just learn about AI security — you’ll build it yourself.
🔐 What You’ll Learn
• How jailbreak attacks work and why they’re dangerous
• How to implement safety guardrails using Google ADK
• Applying LLM-as-a-Judge to detect unsafe outputs
• Using Model Armor to automatically block harmful behavior
• Designing secure and stable multi-agent systems
🎤 Speaker: Khanh Linh Nguyen
Head of AI, Obello
Google Developer Expert, AI/ML
📅 4 March 2026 (Wednesday)
⏰ 4:00 PM – 5:30 PM
🌐 Fully Virtual
Building agents is powerful.
Building secure agents is professional.
📝 Register Now
🔗 https://forms.gle/QvYaYkaCyagvYSCTA
📣 Stay Connected
Telegram: https://t.me/gdgutm
Instagram: https://www.instagram.com/gdg.utm/
LinkedIn: https://www.linkedin.com/company/gdgutm/
Gemini Nexus: The Agentverse
Build with AI 2026
Organized by
Google Developer Groups On Campus
Universiti Teknologi Malaysia
Strategic Partner
Google Developer Group George Town
#GeminiNexus #Agentverse #GDGOnCampus #GDGGeorgeTown #BuildWithAI #AIAgents #GoogleAI #Malaysia
Google Developer Expert AI/ML, Head of AI @ Obello (Silicon Valley)
