Secure Code By Design

This is part of my blog series on Secure Code By Design .

I originally posted this on Medium and moving it over to my substack blog into the Learn Security  section of my Substack security and technology research blog . The revised posts will be reviewed and updated to add anything that may be missing related to AI-generated software.

From what I’ve seen to date, AI models are not writing software free from vulnerabilities. And for various reasons, they are still going to need guidance. Even as they improve they can randomly change and cause problems as I explained in this post on Anthropic Mythos . Understanding these concepts will help you write more well-architected and secure AI generated code as well as hand written code. As always the tools are helpful but understanding the underlying principles will help you become more proficient and valuable as a software engineer, software architect, or security engineer in the long run - not just a vibe coder.

Similar to my last book, Cybersecurity for Executives in the Age of Cloud , I started writing this new book on secure coding in the form of blog posts. This post is simply a place to list all the related topics I’ve written so far in one place. I’ve decided not to publish an actual book and will just be completing this book as time allows on my blog.

The book is focused on code. How can you write defensive software that reduces the chances you give attackers to break into systems? By incorporating some of the concepts I’ll write about into your programming practices you should be able to reduce the risk of software vulnerabilities in your applications at the core, and more holistically.

Have a question about secure software programming or software development using AI? Founding members on my substack blog can ask me questions. Subscribe for new posts on this topic (i.e. chapters in this book) as they are updated and added.

You can read the full post here:

— Teri Radichel