AI in the SDLC: Separating the Hype from the Security Reality

Autonomous AI development isn't coming, it's already here. McKinsey found that 31% of enterprises now run at least one AI agent in production, many in high-stakes environments like banking, financial services, and healthcare.

Implementation of autonomous AI agents is outpacing regulation, and the risk signals are alarming. Secure Code Warrior and RMIT University research found no universally secure AI coding model. AI-generated code carries an average of 15 vulnerabilities per codebase. Every developer, even "citizen developers," needs to understand the risks before code reaches production.

Secure Code Warrior and KnowBe4 are teaming up to unpack the market shifts driving this change and the vulnerability patterns showing up in AI generated code. Earlier this year, the two partnered to bring secure coding content into KnowBe4's attack simulation and training library. In this live webinar, they’ll share practical techniques for reducing AI development risk.

KnowBe4 CISO Advisory Kawin Boonyapredeeto will discuss software development's evolution from compilers and CI to today's goal-driven AI agents. He'll cover:

• The engineer's job is shifting from writing code to directing it
• Guardrails, specs, and verification are now core engineering skills
• What this shift means for security and engineering leaders 

SCW CEO and Co-Founder Pieter Danhieux will dive into the results of the SCW AI Trust Index, a benchmark of how often leading AI coding models generate vulnerable code. He'll share key highlights from the research, including:

• No model is universally secure: 15 vulnerabilities per codebase, on average
• Vulnerability patterns repeat across auth, input validation, database, and API code
• How to detect AI and MCP influence in your pipeline and reduce the risk 

Don't miss this live session to learn how you can make more secure, cost-effective, and contextually relevant choices when integrating AI tooling into your tech stack.