[Free Community Webinar | Zoom] AI-Native Agent Security & Compliance Part 1: From Chatting to Action, Where Do the Risks Come From?

[Free Community Webinar | Zoom] AI-Native Agent Security & Compliance Part 1: From Chatting to Action, Where Do the Risks Come From?

Following “Embodied Intelligence & Physical AI” and “AI Agents & Neo-Media”, ZJUAANC and NACUAA launch the webinar series: AI-Native Agent Security & Compliance.

AI agents are transitioning from conversational chatbots to autonomous executors. Once connected to tools, files, calendars, codebases, or production environments, the risks transcend generative hallucinations to direct operational impact: unauthorized writes, token leakage, privilege escalation, and unintended tool execution. This series specifically addresses security, access control, and regulatory compliance for AI-native agent products.

Part 1 establishes common ground: how AI agent security fundamentally diverges from traditional network security, what the six most common risk vectors are, and the bare minimum compliance boundaries individuals and small teams must maintain prior to production deployment.

🎥 Direct Zoom Access

📹 Zoom Online Meeting Credentials (Free Public Access)

Topic: AI-Native Agent Security & Compliance · Part 1

Meeting URL: https://zoom.us/j/95095998625?pwd=zbrMxVlR0BInfpYSIwezl31O40kePF.1
Meeting ID: 950 9599 8625
Passcode: 18970521
One-tap mobile: +16699006833,,95095998625#,,,,*18970521# US (San Jose)

🎟️ Register on Luma (https://luma.com/fiht7lab) 👉 Direct Link to Zoom

📅 Date & Time

☀️ Time Correction Notice: The lecture will start at 5:30 PM PDT (previously 6:30 PM).

• California (PDT): Saturday, Sep 5, 5:30 PM – 6:30 PM
• Eastern (EDT): Saturday, Sep 5, 8:30 PM – 9:30 PM
• Beijing (CST): Sunday, Sep 6, 8:30 AM – 9:30 AM

📋 Lecture Outline

1. Chatbots vs. Autonomous Agents: Where the attack surface fundamentally shifts
2. Six High-Incidence Risk Vectors: Excessive permissions, API key leakage, indirect prompt injection, destructive tool misuse, unauthorized data retrieval, and unclear accountability
3. Pre-Deployment Minimum Compliance Checklist for individuals and bootstrap startups
4. Curriculum Roadmap: Previews of Parts 2–5 and open Q&A

🎯 Target Audience

Beginners seeking AI literacy, builders and engineers deploying agents in practice, and founders/creators concerned with production governance and compliance.

📝 Community & Registration

• WeChat Community: Connect with zjuaanc (Note: Agent Security).
• Luma Registration: https://luma.com/fiht7lab

💡 Free public event. We will never ask for wire transfers or verification codes.

🤝 Organizers & Collaborators

NACUAA AI Venture & Impact Club, ZJUAANC · ZJU AI Agents, VI AI Foundation, VI AI for Good Foundation, in collaboration with AI Agent Alliance

📰 Media Partner

North Alpha Post