Google, under Sundar Pichai’s leadership, is broadening its AI initiatives by launching CodeMender, an AI security tool designed to identify and fix vulnerabilities in software code.
Sundar Pichai-led Google is expanding its AI ambitions beyond chatbots and Search. The Alphabet-owned search giant has now introduced CodeMender, a new AI security agent that can automatically detect vulnerabilities in code, suggest fixes, test them, and even deploy patches with human approval.
Google announced CodeMender to compete with Claude Mythos, a new AI tool from Anthropic created to help companies find and fix security issues in their software.
“Leveraging Agent Platform capabilities and advanced Gemini models, CodeMender autonomously identifies vulnerabilities within your code. It then recommends precise fixes, securely tests them, and can apply patches and necessary changes across dependent systems, with your approval. This entire process automates secure deployment while ensuring your developers retain control,” the company said in a blog post.
CodeMender was originally developed by Google DeepMind and is now being integrated into Google’s Agent Platform. The tool runs on advanced Gemini models, including Gemini 3.5, and is designed to help companies secure their software faster.
For those unaware, CodeMender works like an AI security engineer. It scans source code to find bugs and security loopholes, recommends exact fixes, tests whether the fixes work and can apply those changes across connected systems after getting approval from developers.
Google says this approach can help businesses reduce the time it takes to patch vulnerabilities while keeping engineers in control of the final decision. Several Gemini Enterprise customers are already testing CodeMender, though Google has not announced a wider release date yet.