Google's CodeMender steps up to rival Anthropic's Mythos in AI security

Google is making a major push into AI-driven cybersecurity. At its I/O event, the company announced it's inviting select groups of experts to test the API for CodeMender, an AI agent for code security first released last October. The tool not only flags vulnerabilities but also fixes them, positioning it as a direct competitor to Anthropic's headline-making Mythos Preview. Google DeepMind CTO Koray Kavukcuoglu told The Verge that the goal is to "help secure the world's code bases," and Google has already initiated talks with governments and enterprises about using CodeMender for system audits.

The timing is deliberate. Anthropic's surprise Claude Mythos Preview, deemed too powerful for public release, sent shockwaves through the AI industry and attracted interest from top banks and even the Federal Reserve chair. That model's supposed capability to uncover unknown security gaps in high-stakes systems prompted a wave of imitators. OpenAI quickly followed with its own offering, and now Google is leveraging CodeMender to stay competitive. CEO Sundar Pichai acknowledged Mythos's impact, stating, "What Mythos has done, and credit to them, is to show that there is a value for the largest-sized model in these kinds of security use cases. But I think it's something we are capable of doing as well." As AI labs face pressure to monetize before IPOs, cybersecurity is emerging as a key revenue driver.