GitHub's AI Coding Boom Exposes a Bigger Security Problem

GitHub's AI coding boom is exposing a major problem in software security. As AI coding agents generate more code and developers push commits faster than ever, can traditional DevSecOps and shift-left security practices keep up?

In this episode of ShipTalk, cybersecurity leader Jake Bernardes breaks down GitHub Universe's AI coding trends and argues that shift-left security is dead in the age of AI-generated code.

Adam Arellano and Martin Reynolds challenge that argument, exploring how AI-powered security testing, automated software delivery, and modern DevSecOps practices can help organizations maintain security without slowing down development.

Topics covered:

GitHub Universe and the explosion of AI-assisted coding

Why AI-generated code creates new software security risks

Is shift-left security still effective?

The growing gap between code generation and security testing

AI coding agents, vulnerabilities, and software supply chain security

How Harness is approaching secure, end-to-end software delivery with AI-powered automation and guardrails

The bigger question: AI coding tools are helping developers write software faster than ever. But are we actually building better, more secure software—or just creating vulnerabilities faster?

Learn more about Harness and AI-powered software delivery: https://www.harness.io

#GitHub #AICoding #DevSecOps #SoftwareSecurity #AIAgents