Cisco Praises Anthropic's Mythos Without Bug Count

Cisco highlights Anthropic's Claude Mythos Preview and OpenAI's GPT 5.5-Cyber as AI tools that scanned 1.8 billion lines of code in eight days. Cisco SVP Anthony Grieco commends the models in a blog post but does not disclose how many vulnerabilities were detected. The scan was conducted over eight days and covered 1.8 billion lines of code. The evaluation focuses on AI's ability to detect code flaws at scale. Cisco's praise does not include specific numbers of bugs found or remediated.