Announcing MCP Toolbox Java SDK v1.0: Agentic data access for the enterprise
cloud.google.com·6d ago
Cloud & Infrastructure
The New Stack·5d ago·5 min read
Claude’s cyber safeguards are getting more flexible — but not for everyone.
TL;DR
Anthropic has restructured its Cyber Verification Program (CVP) into three tiers, allowing varying access to its AI model Claude Opus 5.5. However, many security teams will still face significant restrictions, particularly in the most limited Defense Access tier.
✦ Why It Matters
Evaluate your team's eligibility for CVP tiers to optimize access to AI capabilities.
Key Takeaways
- Assess your team's fit for CVP's three access tiers.
- Prepare for potential restrictions in the Defense Access tier.
- Consider applying for Red Team Access for more flexibility.
- Stay informed about upcoming data storage options with Enterprise Frontier Safeguards.
How It Works
The CVP's three tiers are designed to match the capabilities of Claude Opus 5.5 with the specific needs of security teams. Each tier has different verification requirements, allowing teams to access the model's capabilities that align with their cyber work.
⚠ The Catch
Many security teams will still face significant restrictions in the Defense Access tier, which may hinder their ability to perform essential defensive tasks.