Trust, but Verify: Securing AI-Assisted Software Development

Dinesh Bhattarai

AI can generate code in seconds, but secure software requires far more than a good prompt.

This session focuses on what comes after code generation: validating AI outputs, securing prompts and development workflows, identifying common security pitfalls in AI-generated code, and building confidence through code reviews, automated testing, and modern security practices.

Rather than treating AI as an infallible developer, we’ll explore how to use it as a powerful engineering assistant while keeping humans in control of architecture, security, and quality. Through practical examples and real-world workflows, you’ll learn how to build software that is not only faster to develop, but also secure, maintainable, and ready for production.

Whether you’re building WordPress plugins, themes, or web applications, you’ll leave with practical techniques to integrate AI into your development process without compromising security or reliability.