This practitioner‑focused programme provides a comprehensive introduction to how agentic AI systems are built, their security risks, and how they can be effectively secured. Led by Dr Kang Li (KangGuru.ai), the course covers real-world security risks, alongside practical defence strategies. Through hands-on labs and industry‑aligned frameworks, participants will gain the skills to identify risks, implement defence‑in‑depth security, and build secure, compliant AI agent systems for real‑world deployment.

Learning outcome

- Understand how AI agents work and how they are built

- Understand the Secure by Design principles, and identify common security risks in agentic AI systems

- Learn how attackers exploit AI agents and how to prevent these attacks

- Apply key security controls such as sandboxing, guardrails, and credential custodian

- Develop practical skills to secure AI systems and manage risks in real‑world use

Target Audience

- AI professionals and engineers who want to understand and secure AI agent systems

- Startup founders and product teams building or using AI agents in real‑world applications

- Analysts and platform teams involved in deploying and managing AI or machine learning systems

- Senior leaders and decision‑makers responsible for AI strategy, risk, and compliance

- Risk and security professionals looking to strengthen AI governance and threat management

*Participants shall bring their own laptop and stationery for the classwork

Course Details

Date: 21 July 2026 (Tuesday)

Time: 1:15pm – 5:15pm 

Language: English

Venue: Multi-Function Room, Smart-Space FinTech, Level 4, Core E, Cyberport 3

Training Fee: 
Refundable commitment fee to Cyberport community members (i.e. Cyberport incubatees, CCMF grantees, Cyberport Alumni):

- HK$300 per head
- 100% refund to eligible participants according to Cyberport Academy attendance policy
*Please obtain the discount code by sending email to training@cyberport.hk.

Regular price: 
- HK$840 per head

*Cyberport reserves the right to charge the original price if applicants use a discount code that does not belong to his/her category
*All monies collected will be used for Cyberport’s public mission start-up events and activities

Speaker

Dr Kang Li, KangGuru.ai 

Before working as CTO at CertiK, Dr Kang Li was formerly the Chief Security Scientist and a Senior Director at Baidu Inc. Before that, he served in various leadership and consulting roles at cyber security enterprises and institutes, including the Institute for Cyber Security and Privacy at UGA, Intel, and Samsung. Dr Kang Li was the founding mentor of multiple CTF teams, inculding Blue-Lotus and Team Disekt. He is a frequent speaker at major security conferences, and he is also a DARPA Cyber Grand Challenge Finalist.

For further training enquiry, please contact Ms. Hazel Tsang at training@cyberport.hk