Anthropic launches Claude Security public beta for enterprise customers, using Opus 4.7 to detect and validate vulnerabilities in production code, with automated patches and noise reduction features.
Smart Scanning for Production Code
Anthropic has made its AI-driven security tool, Claude Security, available in public beta for Claude Enterprise customers. The tool uses the Opus 4.7 model to find security flaws in production code without requiring extra tools or API connections. It performs full security checks, validates each finding to minimize false alarms, and suggests patches for developers to review before deploying changes.
Enhanced Features from User Feedback
The tool first appeared as a research preview in February 2026 and has since been used by hundreds of organizations to uncover vulnerabilities missed by other scanners. Based on early feedback, Anthropic added scheduled scans for regular checks, directory-level targeting to focus on specific code sections, CSV and Markdown export options, webhook notifications for instant alerts, and persistent dismissals that reduce noise over time. The new validation feature tackles a common complaint about automated scanners: too many false positives leading security teams to ignore alerts. Claude Security aims for a higher signal-to-noise ratio than traditional static analysis tools.
Straightforward Entry for Security Teams
Anthropic says the tool is designed as a direct on-ramp with no need for custom agents or API setup. For enterprise security teams wanting to expand vulnerability coverage without hiring more staff or building internal AI systems, Claude Security offers a low-barrier starting point.
Source: Cyber Security News
