Anthropic launches Claude Security to scan code for flaws and prioritize fixes for enterprises

Anthropic Brings Claude Security to Enterprise Customers

Anthropic has announced the public beta launch of Claude Security, an AI-powered cybersecurity tool designed to help enterprise security teams scan codebases for vulnerabilities and generate targeted patches1

1

. The tool is now available to Claude Enterprise customers globally, with access for Team and Max-tier users coming soon2

2

. Originally introduced as Claude Code Security in research preview in February, hundreds of organizations have already used it to discover and fix exploits in production code, including software vulnerabilities that existing tools had missed for years4

4

.

How Claude Opus 4.7 Model Powers Vulnerability Detection

Claude Security is powered by the Claude Opus 4.7 model, which Anthropic describes as "among the strongest models available for finding and patching software vulnerabilities"5

5

. Unlike traditional scanners that match against known vulnerability patterns, the tool reasons about code the way a security researcher does, tracing data flows, reading source code, and examining how components interact across files and modules1

1

. This approach enables the system to discover complex, context-dependent issues that might otherwise be missed. A multi-stage validation pipeline verifies each finding independently before it reaches an analyst, with every result assigned a confidence rating and severity rating4

4

.

From Scan to Fix: Accelerating Remediation for Cybersecurity Defenders

Anthropic emphasizes that "time from scan to fix is the metric that matters," with early users reporting they can go from scan to applied patch in a single sitting instead of days of back-and-forth between security and engineering teams5

5

. Once Claude Security identifies an issue, users can move directly into a Claude Code session to review and apply a fix with full context on the issue's likely impact and reproduction steps4

4

. The tool generates patches and fixes that enable organizations to find and fix software vulnerabilities at scale2

2

. After feedback from the research preview, Anthropic added scheduled scans for ongoing coverage, the ability to dismiss findings with documented reasons, and comma-separated and Markdown exports for easy import into existing audit systems2

2

.

Enterprise Partnerships and Platform Integrations

Service partners including Accenture, Deloitte, Infosys, and PwC are already deploying Claude to ramp up enterprise security3

3

. Harpreet Sidhu, global lead of Accenture Cybersecurity, stated that "to combat AI-fueled cyber threats, our clients need advanced solutions they can deploy today, not tomorrow"3

3

. Technology partners including CrowdStrike, Palo Alto Networks, SentinelOne, Trend.ai, Wiz, and Microsoft are integrating Opus 4.7 into their cybersecurity platforms2

2

5

5

.

Broader Cyberdefense Strategy and Security Safeguards

Claude Security sits alongside Anthropic's broader cyberdefense initiatives, including Project Glasswing, which brings together technology partners to secure the world's production software using the Mythos model2

2

. However, Claude Security does not leverage the Mythos model, which remains available only to Project Glasswing participants5

5

. To prevent misuse, Opus 4.7 includes new cyber safeguards that automatically detect and block requests suggestive of prohibited or high-risk cybersecurity uses, such as mass data exfiltration or ransomware code development1

1

. The model also blocks activities with legitimate defensive applications, though cybersecurity researchers approved to join Anthropic's Cyber Verification Program gain access to AI capabilities in this restricted area1

1

. This dual approach aims to make frontier AI capabilities available to defenders while limiting access for potential attackers.