Code Security Reinvented: Navigating the Era of AI
summary
GitHub Security Lab reports 1,000+ issues found in open source with 600+ CVEs and a 96% fix rate; demos show using GitHub Copilot to write safer code, run CodeQL queries, and generate fuzzing inputs. Secret scanning blocks ~30,000 secrets in 2023, Copilot suggestions account for ~60% of committed code (up from 35%), and a code scanner autofix (public beta) proposes multi-file PR fixes. A free training game has 2,000 stars and 6,000 active players.
AI-written summary · about
chapters (17)
- 0:00AI-driven code security intro
- 1:06Security expertise gap
- 3:57AI security use cases
- 4:42Safer code with AI context
- 8:19AI security limitations
- 11:12AI secret scanning
- 12:39AI-assisted CodeQL
- 14:29AI automated security fixes
- 16:16AI security training
- 18:22AI vulnerability patterns
- 20:32AI security guidance
- 22:20AI fuzzing for security
- 24:06AI privacy handling
- 25:33Developer priorities survey
- 28:06AI security integration summary
- 29:09Interactive security game
- 30:54Q&A
related talks
38mCode Security Reinvented: Navigating the era of AI
41mCode Security Reinvented Navigating the era of AI
58mSecuring Code in the Age of AI
30mHuman vs AI: How to Ship Secure Code
23mCode Security Reinvented Navigating the era of AI
26mSmarter Code, Sneakier Risks: Supply Chain Security in the Age of AI
52mSecurity as Code: A DevSecOps Approach
19mFlipping the script on application security