Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstalliTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Anthropic’s Cyber Verification Program (CVP) is a tiered, application-based access program for verified security professionals. It can provide less restrictive cyber safeguards on selected Claude models, but it is not unrestricted offensive access and approval is not guaranteed. Defense Access covers defensive security work, Red Team Access adds authorized adversarial testing for organizations, and Specialized Access is a tightly reviewed tier for organizations testing safety-sensitive systems such as power grids or flight systems.
Anthropic announced the expanded structure on October 6, 2026. Applicants should expect identity and organization checks, evidence of security controls, continuing authorization limits, and terms that vary by model and cloud platform.
What is Anthropic’s Cyber Verification Program?
CVP is Anthropic’s controlled route for giving qualified cybersecurity practitioners access to advanced cyber capabilities while managing dual-use risk. Anthropic describes the problem directly: “Cybersecurity is inherently dual use: the same capabilities that enable a security team to find and fix a vulnerability can also help a malicious actor exploit it.”
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsThe expanded program integrates Project Glasswing, an earlier defensive initiative, into three access tiers. Anthropic lists Claude Opus 5.5, Claude Sonnet 5.5, Claude Mythos 5.1 and future models among the models covered by the announcement. Model eligibility, controls and availability can change, so an applicant must confirm the terms for the specific model and platform being requested.
#1 Best Overall
CVP is not a general-purpose permission slip for hacking. Each tier is tied to a defined scope of work, and safety blocks remain active for activities that could cause physical harm or mass disruption.
Which CVP tier fits your work?
| Tier | Permitted focus | Who may apply | Review expectation | Important boundary |
|---|---|---|---|---|
| Defense Access | Security operations, incident response, malware reverse-engineering, vulnerability analysis and validation | Organizations and, in suitable cases, individual researchers with a record of reported vulnerabilities | Anthropic says many defensive teams should qualify and aims to respond within a few days; this is an expectation, not a promise | Defensive purpose and ordinary authorization requirements still apply |
| Red Team Access | Everything in Defense Access plus authorized penetration testing and red-teaming | Organizations only; individuals are currently excluded | Anthropic expects a few weeks because eligibility and controls are more demanding | Testing is limited to systems the organization is authorized to test |
| Specialized Access | Intensive testing of safety systems whose failure could affect lives or disrupt markets | A limited set of verified organizations | In-depth, case-by-case review with collaboration from the U.S. government | Designed for high-consequence systems, not ordinary application testing |
Defense Access
Defense Access is the entry point for most legitimate defensive use. Examples include a company security team investigating an intrusion, a university analyzing malware, an incident-response provider validating a vulnerability, or an open-source maintainer checking a patch. Anthropic also gives examples such as nonprofits and government bodies defending systems they own or maintain, regional hospitals and municipal utilities, smaller security firms, and individual researchers who have a demonstrated history of reporting vulnerabilities.
The examples describe the kind of work Anthropic intends to support; they do not create an automatic entitlement. Every applicant is still verified, and Anthropic can decide that a particular use, organization or control environment does not fit.
Free tools Windows power users keep installed
One-click scans. No signup required.
Red Team Access
Red Team Access extends the defensive scope to authorized adversarial testing. Anthropic names in-house and government red teams and security or penetration-testing firms as examples. A qualifying organization is enrolled in Defense Access while its Red Team request is reviewed.
The authorization requirement is central. Anthropic’s program description says organizations in this tier can perform adversarial testing only against systems they are authorized to test, including IT systems in critical industries. A security consultant cannot use approval to probe an unrelated company, public service or cloud tenant without documented permission. Individual researchers cannot currently apply for this tier, even if they have a strong vulnerability-disclosure history.
Specialized Access
Specialized Access is reserved for a small number of verified organizations that are authorized to test safety-sensitive systems. Anthropic’s examples include flight operating systems, power grids, telecommunications networks, interbank-transfer infrastructure and government administrative networks.
Because an error in these environments could endanger lives or destabilize markets, Anthropic says it reviews each organization in depth and works with the U.S. government. This is not simply a higher usage quota or a faster version of Red Team Access; it is a distinct authorization and risk-review process. Existing Project Glasswing members transition to Specialized Access for current models without reapproval, according to Anthropic’s October announcement.
Who can apply?
CVP is aimed at verified professionals and organizations with a legitimate security mission. The applicant categories Anthropic identifies include:
- Security teams at companies, nonprofits, universities and government bodies defending systems they own or maintain.
- Operators of critical infrastructure, including regional healthcare providers and municipal utilities.
- Security, incident-response and penetration-testing firms.
- In-house and government red teams.
- Open-source maintainers responsible for software used by others.
- Individual security researchers with a track record of reporting vulnerabilities, for defensive work under Defense Access.
These categories are examples rather than a published acceptance guarantee. Anthropic says it will verify every applicant and request evidence of the security controls required for the requested tier. The appropriate application therefore needs to explain your organization, the systems you maintain or are authorized to test, the people who will use the service, and how access is controlled.
Can individual security researchers apply?
Yes, potentially for Defense Access. Anthropic specifically includes individual researchers with a record of reported vulnerabilities among its examples. The record matters: a stated history of responsible disclosure is more relevant than a general claim of interest in security.
Individuals cannot currently apply for Red Team Access. That tier is organization-only because Anthropic expects formal authorization, governance and security controls around adversarial testing. Specialized Access is also an organizational tier.
What happens during verification and review?
Anthropic does not publish a permanent, universal checklist in the October announcement. Requirements can change by tier, model and platform. Its stated process has several practical implications:
- Select the intended tier and use case. Describe whether the work is defensive analysis, authorized red teaming or testing of a safety-sensitive system. Requesting a broader tier than your authorization supports can create unnecessary review friction.
- Provide identity and organization information. Anthropic says every applicant is verified. Organizations should be prepared to establish who they are, what systems they operate and who will administer access.
- Show the relevant security controls. Anthropic says it will request proof of controls required for the tier. Do not assume that a signed statement alone substitutes for the evidence requested in the current application.
- Document authorization. Red-team work must target systems the organization is authorized to test. Keep scope, ownership and testing permission clear enough for reviewers and administrators to enforce.
- Plan for a review window. Anthropic aims to respond to many Defense Access applicants within a few days, while it expects Red Team Access review to take a few weeks. These are stated expectations, not service-level commitments.
- Assign approved access to workspaces. Administrators, rather than every member of an organization, assign access to specific workspaces. Confirm who has that administrative responsibility before requesting enrollment.
Approval should be treated as conditional access. Anthropic can maintain blocks, require ongoing monitoring and change model or platform eligibility after enrollment.
What participants can and cannot do after approval
Safety restrictions remain active
CVP reduces some cyber-related restrictions for verified users; it does not remove all safeguards. Anthropic says users can still encounter real-time blocks for actions that could cause physical harm or mass disruption. Examples include deploying ransomware, damaging physical systems and penetration-testing high-risk safety systems outside the permissions of the applicable tier.
Red Team Access therefore does not mean “test anything.” The organization must have authorization for the target, and the request must remain within Anthropic’s safety boundaries. Specialized Access is designed for approved safety-system testing, but even that tier is subject to its own review and controls.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Workspace governance matters
Because administrators assign access to specific workspaces, organizations should separate approved cyber work from general experimentation. Limit membership to personnel with a business need, keep target authorization records with the project and establish a process for removing access when a contract or engagement ends.
Data retention and platform availability
Retention terms are a material part of the decision. Anthropic says retention is required for enrolled organizations so it can monitor for cyber misuse. The same October announcement describes zero-data-retention cases for organizations using Claude Fable 5.1 or Claude Mythos 5.1 with zero data retention, and says Enterprise Frontier Safeguards (EFS), which combines zero data retention with safeguards, was expected later in fall 2026.
Those statements do not create a universal zero-retention promise. Anthropic’s separate Mythos information page says Mythos 5.1 uses a 30-day data-retention policy by default. The applicable arrangement depends on the model, contract, platform and whether the organization qualifies for the relevant EFS or zero-data-retention configuration. Confirm the current terms in writing before sending sensitive logs, malware samples or proprietary source code.
The October 6 announcement lists CVP availability on the Claude Platform, Google Cloud Vertex AI and Microsoft Foundry. Amazon Bedrock access is limited to customers eligible for EFS. Cloud availability and model support are volatile; an organization should check the current CVP and platform documentation at the time of application rather than relying on an older deployment plan.
What Anthropic reports about results
Anthropic published program figures, but they are company-reported measurements rather than independent estimates of what every participant will achieve.
| Reported result | What it represents | Qualification |
|---|---|---|
| At least 129,000 verified software vulnerabilities | Partner results between April and July 2026 | Anthropic’s lower-bound figure from partial partner data |
| 5,500 additional verified vulnerabilities | Anthropic open-source scanning between April and October 2026 | Reported by Anthropic; methods and disclosure rates varied |
| More than 33,000 high- or critical-severity vulnerabilities | Severity subset of reported verified findings | Anthropic calls this likely an undercount and based it on partial data from 33 partner reports |
| 46 of 50 trials blocked | Claude Opus 5.5 in Defense Access across five attempts at each of ten CyScenarioBench challenges | Measures safeguard behavior in an evaluation, not real-world prevention |
| 34 of 50 tasks completed | Claude Opus 5.5 in Red Team Access on the same evaluation | Anthropic reports a 67.6% success rate with no safeguards applied |
For the vulnerability totals, Anthropic says the numbers are a lower bound because they use partial data from 33 partner reports and open-source partnerships. Organizations used different triage approaches, and fewer than half disclosed how many findings had been patched, often because remediation was still underway. Anthropic projects that the true impact could be at least five times higher; that is Anthropic’s projection, not an independently established count.
The CyScenarioBench figures also need careful reading. Anthropic says every task was blocked on the first prompt without CVP access, 46 of 50 trials were blocked at some point in Defense Access, and Red Team Access produced no blocks and 34 successful completions out of 50. There were five trials for each of ten challenges; the 50 trials are not 50 distinct vulnerabilities. The evaluation concerns safeguards on complex interactive offensive scenarios and does not guarantee a particular result in a real engagement.
How Project Glasswing relates to CVP
Project Glasswing was announced on April 7, 2026 as an effort to secure critical software with Claude Mythos Preview. Anthropic named Amazon Web Services, Anthropic, Apple, Broadcom, Cisco, CrowdStrike, Google, JPMorganChase, the Linux Foundation, Microsoft, NVIDIA and Palo Alto Networks as launch partners, and said access had been extended to more than 40 additional organizations building or maintaining critical software infrastructure.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →Under the October CVP expansion, Glasswing is being integrated into the broader program and existing members move to Specialized Access for current models. A Glasswing partnership or launch-partner name does not establish that a new applicant is endorsed, automatically eligible or entitled to the same terms.
Programs that are not CVP
Model Safety Bug Bounty Program
Anthropic’s Model Safety Bug Bounty Program is a separate, invitation- and acceptance-based program for selected participants to report specified model jailbreaks. It has its own application process and confidentiality obligations. Participation there does not answer whether an applicant qualifies for CVP.
Claude Security
Claude Security is a separate product described by Anthropic as scanning codebases for vulnerabilities and suggesting targeted patches for human review. Using or evaluating that product is not the same as enrolling an organization in CVP.
A practical decision checklist before applying
- Define the work: defensive operations, authorized red teaming or safety-system testing.
- Confirm legal authority: identify the owner and written scope for every system that may be tested.
- Choose the narrowest suitable tier: do not request Specialized Access for ordinary vulnerability validation.
- Prepare verification evidence: organization identity, responsible administrators, user groups and the security controls relevant to the tier.
- Plan retention: determine whether the selected model and platform require retention, support zero data retention or fall under EFS.
- Check deployment support: verify model and platform availability for Claude Platform, Vertex AI, Microsoft Foundry or, where eligible, Amazon Bedrock.
- Set operating controls: restrict workspace membership, retain authorization records and define incident escalation and offboarding.
- Set expectations: an application can be declined, review times are estimates and safety blocks continue after approval.
Bottom line for prospective applicants
CVP is best understood as verified, scoped access for security work—not unrestricted access to offensive capabilities. Defense Access can accommodate organizations and some individual vulnerability researchers. Red Team Access is for organizations conducting authorized adversarial testing, and Specialized Access is for a limited group testing systems whose failure could affect lives or markets. Verification, security-control evidence, authorization boundaries, retention terms and platform eligibility all matter as much as the model name. Check Anthropic’s current CVP and Help Center materials immediately before applying because requirements, review timing, supported models and data-handling options can change.
Recommended Free Tools
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
