Google wants to compete with Anthropic’s Mythos

Google is positioning its "CodeMender" AI agent to compete with Anthropic’s high-profile "Mythos" security model. Originally introduced last October, CodeMender is now being opened to select experts for API testing, with Google marketing it as a solution to detect and resolve vulnerabilities in global code bases. This move comes as major AI firms, including OpenAI, increasingly view cybersecurity as a lucrative market for government and enterprise clients. Google DeepMind CTO Koray Kavukcuoglu noted that the company is already in discussions with high-stakes organizations, while CEO Sundar Pichai emphasized that Google possesses the technical capability to match the security performance demonstrated by Anthropic’s model.

(Source：The Verge)