Google has announced CodeMender, a new AI-powered system that detects and fixes software vulnerabilities automatically in real time.
Powered by Gemini Deep Think models, CodeMender utilizes advanced reasoning, debugging, and validation methods to identify code weaknesses and apply high-quality security patches that resist future exploits.
Within the past six months alone, the platform has contributed more than 70 fixes to major open-source projects—some with millions of lines of code —through upstreaming.
Through automation of patching and validation of security, CodeMender liberates developers to build innovation instead of investing time on manual vulnerability management.
Aside from reactive patches, CodeMender can pre-rewrite insecure code by adding safer data structures and APIs, securing software against future exploits.
In one instance, the AI enhanced a top image compression library by introducing compiler-level annotations that prevent buffer overflow attacks—basically disabling exposures like those used in real exploits.
Although all CodeMender patches are currently human-reviewed, Google plans to extend this review process to core open-source libraries and eventually open-source the technology for developers worldwide.
You may also want to check out some of our other recent updates.
Wanna know what’s trending online every day? Subscribe to Vavoza Insider to access the latest business and marketing insights, news, and trends daily with unmatched speed and conciseness! 🗞️
