Anthropic announced an expanded version of its Cyber Verification Program (CVP), which makes advanced cyber capabilities and reduced blocking classifiers available to qualifying security professionals, according to a post on the company's website dated October 6, 2026. The program now consists of three access tiers, each including access to models such as Claude Opus 5.5, Claude Sonnet 5.5, and Claude Mythos 5.1. Anthropic says its generally available models carry conservative cyber safeguards that block most cyber work because cybersecurity is 'inherently dual use.' The expansion integrates two prior programs: Project Glasswing, which gave access to Claude Mythos, and the earlier CVP, which gave vetted teams reduced safeguards on Claude Opus and Sonnet. The new tiers are Defense Access (for defensive work such as incident response and malware reverse-engineering), Red Team Access (adding authorized penetration testing, organizations only), and Specialized Access (for a limited set of organizations authorized to test safety-critical systems like power grids and flight operating systems, reviewed in collaboration with the US government). Anthropic states it expects to respond to Defense Access applications within a few days and Red Team applications within a few weeks. Data retention is required for enrolled organizations to monitor for misuse, with a zero-data-retention option planned via Enterprise Frontier Safeguards later in the fall.
- Three access tiers: Defense Access, Red Team Access, and Specialized Access, each with different verification and security controls.
- Consolidates two prior programs (Project Glasswing and the earlier CVP); existing Glasswing members move to Specialized Access without reapproval.
- Data retention is required for enrolled organizations, with a zero-data-retention option (Enterprise Frontier Safeguards) planned for later this fall.
- Specialized Access organizations are reviewed in depth in collaboration with the US government.
What it means for you
This matters only if your organization does professional security work — defending systems you own, running a security operations center, or doing authorized penetration testing. Anthropic's standard models deliberately block most cyber tasks, and this program is the formal route for vetted teams to get those blocks relaxed. For everyone else, it changes nothing about how you use Claude day to day.
Try this
If you have a security or IT team doing defensive work, have whoever owns that function check the Defense Access eligibility list to see whether applying is worth their time — Anthropic says it aims to respond within a few days.
Who should care
In-house security teams, SOC and incident-response staff, penetration testing firms, open-source maintainers, and operators of critical infrastructure such as regional hospitals or municipal utilities.
Skip this if
Your organization does not do dedicated cybersecurity work, or your security needs are limited to code review, patching, and alert triage — which the generally available models still handle.
Sources: Anthropic — read the original