Unbreaking Access Control with AI Coding Agents?

Can AI coding agents help us fix access control—or are they just making things worse?

Unbreaking Access Control with AI Coding Agents ?

Broken access control continues to be one of the most critical and pervasive vulnerabilities in modern web and application security. Despite decades of awareness, it remains at the top of incident reports and real-world breaches.

In this interactive workshop, we explore a provocative question:
Can AI coding assistants help us fix access control—or are they just making things worse?

What to expect

• A concise overview of why access control keeps breaking
• Real-world examples of common and subtle authorization flaws
• Hands-on insights into how AI-assisted development impacts security
• Demonstrations showing how AI tools generate (in)secure access control logic

Who should attend

• Software engineers and architects
• Security engineers and AppSec professionals
• Researchers and students interested in secure development
• Anyone curious about the intersection of AI and software security