Thursday
Room 1
09:00 - 10:00
(UTC+02)
Talk (60 min)
Generating Secure C Programs with AI
This presentation examines the integration of AI agents into the secure C programming lifecycle.
Through an iterative case study—generating a consensus-determination program—it demonstrates how refining prompts and enforcing modern C23 standards improves code safety and compliance. The discussion highlights the dual nature of AI in C development: it significantly accelerates scaffolding, testing, and standardization but necessitates expert human oversight to identify sneakily wrong code, avoid undefined behavior, and ensure optimal utilization of modern language features.
Robert Seacord
Robert C. Seacord is a Senior Staff Software Engineer at Woven by Toyota where he works on the Software Craft. Robert is convener of the ISO/IEC JTC1/SC22/WG14 international standardization working group for the C programming language. He is the author of seven books, including Effective C, Second Edition (No Starch Press, 2024), The CERT C Coding Standard, Second Edition (Addison-Wesley, 2014) Secure Coding in C and C++, Second Edition (Addison-Wesley, 2013), and Java Coding Guidelines: 75 Recommendations for Reliable and Secure Programs (Addison-Wesley, 2014). He has also published more than 50 papers on software security, component-based software engineering, Web-based system design, legacy-system modernization, component repositories and search engines, and user interface design and development. Robert has been teaching secure coding in C and C++ to private industry, academia, and government since 2005. He started programming professionally for IBM in 1982.