The One About AI Hacking & Control
As AI becomes increasingly integrated into applications and workflows, it is creating new opportunities as well as new security challenges. Join us as we explore how AI systems can be governed, monitored, tested, and secured against emerging threats.
More About The Sharings
Zayne Zhang (Co-Founder & CEO, Hacktron AI) will be sharing on "Hacking AI and Hacking with AI"
AI has not only expanded the attack surface of systems that leverage it but also significantly enhanced threat actors' capabilities. Zayne will explore the critical vulnerabilities they found in Next.js, vinext, Perplexity and Windsurf, and into how Hacktron is approaching AI-assisted security research to uncover previously undiscovered vulnerabilities, even after multiple high-profile penetration tests. (Technical Level: 200)
Benjamin Goh (Head of Forward Deployment, Dynamo AI) will be sharing on “Securing Agentic AI in Production.”
As organisations deploy AI agents, the challenge shifts from experimentation to control. Benjamin will share more on Dynamo AI’s AgentWarden platform, which evaluates, monitors, and secures AI agents against risks such as unauthorised actions, prompt injection, excessive permissions, data leakage, and policy violations. By combining pre-deployment testing, runtime guardrails, and continuous observability, explore how organisations can mitigate the “Lethal Trifecta” and adopt agentic AI safely while maintaining control, compliance, and operational resilience. (Technical Level: 100-200)
Haoyu Wang (AI Scientist, AIDX) will share on "Runtime Safety for Agentic Systems"
As agentic systems take on more complex, real-world tasks, ensuring they behave reliably and safely at runtime becomes a critical challenge. Haoyu will walk through three of his own research papers exploring different facets of agentic safety: a domain-specific language that allows domain experts to specify governance requirements for agents, a learning-based method for proactively forecasting potential risks before they materialise, and a framework for assessing and layering safety into multi-agent coordination. Get a grounded look at what it takes to build agentic systems that are not just capable, but trustworthy. (Technical Level: 200)
More About The Speakers
Zayne Zhang is the Co-Founder and CEO of Hacktron, an AI security engineering platform that helps teams move from pentest request to audit-ready report in hours. Before founding Hacktron, he worked as a software developer and information security professional, and holds a degree in Computer Science from the University of Cambridge. Zayne has been credited for over 15 vulnerabilities in the CVE database and over 50 security issues across bug bounty programmes, and has competed as a finalist at DEF CON 31 and 32 CTF, among other prestigious competitions worldwide.
Benjamin Goh is Head of Forward Deployment and GM for Singapore, Australia, and Southeast Asia at Dynamo AI. His work centers on building the infrastructure for AI reliability — from rigorous evaluations and runtime guardrails to the processes and frameworks that give organizations the confidence to deploy AI at scale. Benjamin holds a Bachelor's degree in Economics and Political Science from New York University and a Master in Public Policy from Harvard University.
Haoyu Wang is a Computer Science PhD candidate at Singapore Management University, where his research focuses on AI security and safety. He is currently an AI Scientist Intern at AIDX, and brings prior experience as a Research Engineer at SMU contributing to work in software testing and programming languages. Haoyu holds a Master's and Bachelor's in Computer Software Engineering from Tianjin University, and has previously interned at ByteDance.
More About The Series
AI Wednesdays is Lorong AI’s weekly gathering, bringing together practitioners, researchers and innovators for technical discussions on research insights, product development and engineering practices.
Get involved: Learn more about Lorong AI | Speaker Sign-up | WhatsApp Community | LinkedIn | X