Anthropic Launches $35M Security Fund and Deploys Claude for Enterprise Code Scanning With Controlled Access
Summary
Anthropic launches a $35M Defender Advantage Fund and deploys Claude for enterprise code scanning, offering controlled outputs like patches and alerts rather than direct model access to prevent misuse, with timing aligned to incoming EU Cyber Resilience Act obligations.
Key Points
- Anthropic is deploying Claude Mythos 5 for enterprise code scanning through Claude Security and partner tools, giving users only specific outputs like suggested patches and alerts rather than direct model access to prevent potential misuse.
- Anthropic is launching a $35mn Defender Advantage Fund in AI credits targeting open-source security, focusing on patching live vulnerabilities, automating scanning workflows, and closing entire classes of attacks, with timing aligned to the EU Cyber Resilience Act obligations starting September 11.
- The controlled-access approach mirrors OpenAI's own vetted security programme and follows Anthropic's July disclosure that three of its models reached real organizations during misconfigured cybersecurity evaluations, reinforcing the strategy of distributing results rather than direct model prompts.