AI-Generated Code: Unmasking the Security Pitfalls
summary
Demonstrates concrete failures and fixes: uses Copilot inside IntelliJ on a Spring Boot CoffeeShop app, GitHub PR checks from Sneak and Burp Suite intercepts to show a SQL injection detected in generated query and a path traversal/overwrite via avatar upload; applies a prepared statement and an AI-proposed patch that removes the vulnerability. Cites tools and numbers—Copilot speeds 50–57% faster, 92% developer adoption, NYU/Stanford finding ~40% of Copilot-generated apps contain vulnerabilities—so the demos prove what actually breaks and what actually fixes.
AI-written summary · about
chapters (10)
- 0:00AI-Generated Code Intro
- 0:47Agenda
- 1:24AI with Security Context Background
- 4:10But What is the Low Hanging Fruit?
- 4:48LLMs: Developer superpower ... security concern
- 7:23Demo: Patch CoffeeShop App
- 8:09Demo: Make product table searchable
- 18:07Demo: Allow users to personalise profiles
- 26:03AI-Generated Code: Takeaways
- 27:00Further Reading
related talks
44mBreaking AI: Live Coding and Hacking Applications with Generative AI
46mBreaking AI: Live coding and hacking applications with Generative AI
30mWhat Are the Security Risks Associated with AI-Generated Code? 🤔
46mBreaking AI: Live coding and hacking applicattions with Generative AI
23mCode Security Reinvented Navigating the era of AI
58mSecuring Code in the Age of AI
41mCode Security Reinvented Navigating the era of AI
31mHacking AI Applications