Anthropic introduced the Life Sciences Verification Program on Sep 17, 2026, giving verified life science teams and institutions access to Mythos, Opus, and Sonnet models under safeguards that are more permissive for biology-related work. The beta program opens applications beyond an early-access group of dozens of organizations and is intended for work such as drug discovery, research biology, clinical development, and manufacturing that may be blocked in generally available Fable models.

Applicants must pass verification covering research credentials, security standards, and ethical research oversight. Verified teams can request Standard Use grants or High-risk Use grants across Claude Science, Claude.ai, Claude Code, and the API. Standard Use grants cover most biology research and development workflows, renew once a year, and currently apply to Mythos 5.1, Opus 5, and Sonnet 5, as well as future models when they launch.

High-risk Use is an add-on grant for work blocked under Standard Use. It removes all safeguards that block life sciences requests, applies to a single research project rather than a full team, and must be renewed every six months. High-risk grants for Claude Opus 5 and Claude Sonnet 5 are available today, while high-risk grants for Claude Mythos remain limited to a small set of entities with additional vetting as Anthropic works with the US government on broader availability.

Anthropic said all other safeguards, including cyber classifiers, remain active under LSVP grants. The company cited increasingly sophisticated misuse attempts on its platform, including activity that could support biological weapons development. LSVP safeguards target three threat models: access compromise through malware or account takeover, insider threats from rogue or coerced employees, and agent misuse involving agents working in swarms or long-horizon tasks.

The program uses a shared-responsibility model in which each entity’s access is tied to the use cases listed in its grant applications. Anthropic will continuously monitor LSVP traffic for usage or patterns outside the stated safe scope and can flag unauthorized activity to organization administrators under pre-agreed timelines for triage and remediation. Use cases should be high-level descriptions, similar to a job listing, and should not include sensitive information or intellectual property.

LSVP shifts some enforcement from real-time blocking to offline monitoring, allowing legitimate work to proceed with fewer interruptions while enabling review of behavior spread across many requests and sessions. Anthropic requires 30 days of data retention for LSVP traffic. The company said the retained data is compartmentalized, cannot be used for model training, and cannot be accessed by Anthropic’s life sciences research teams.

The beta is available in Anthropic’s first-party console for API usage and in Claude for Enterprise and Team plans. Individual plans, third-party platforms, and BAA-enabled organizations are not yet supported, so customers using PHI data should use separate non-BAA organizations with non-HIPAA data. Anthropic expects to enroll hundreds of organizations within the first week and expand support for most of the life science community in the coming weeks.

Anthropic introduced the Life Sciences Verification Program on Sep 17, 2026, giving verified life science teams and institutions access to Mythos, Opus, and Sonnet models under safeguards that are more permissive for biology-related work. The beta program opens applications beyond an early-access group of dozens of organizations and is intended for work such as drug discovery, research biology, clinical development, and manufacturing that may be blocked in generally available Fable models.

Applicants must pass verification covering research credentials, security standards, and ethical research oversight. Verified teams can request Standard Use grants or High-risk Use grants across Claude Science, Claude.ai, Claude Code, and the API. Standard Use grants cover most biology research and development workflows, renew once a year, and currently apply to Mythos 5.1, Opus 5, and Sonnet 5, as well as future models when they launch.

High-risk Use is an add-on grant for work blocked under Standard Use. It removes all safeguards that block life sciences requests, applies to a single research project rather than a full team, and must be renewed every six months. High-risk grants for Claude Opus 5 and Claude Sonnet 5 are available today, while high-risk grants for Claude Mythos remain limited to a small set of entities with additional vetting as Anthropic works with the US government on broader availability.

Anthropic said all other safeguards, including cyber classifiers, remain active under LSVP grants. The company cited increasingly sophisticated misuse attempts on its platform, including activity that could support biological weapons development. LSVP safeguards target three threat models: access compromise through malware or account takeover, insider threats from rogue or coerced employees, and agent misuse involving agents working in swarms or long-horizon tasks.

The program uses a shared-responsibility model in which each entity’s access is tied to the use cases listed in its grant applications. Anthropic will continuously monitor LSVP traffic for usage or patterns outside the stated safe scope and can flag unauthorized activity to organization administrators under pre-agreed timelines for triage and remediation. Use cases should be high-level descriptions, similar to a job listing, and should not include sensitive information or intellectual property.

LSVP shifts some enforcement from real-time blocking to offline monitoring, allowing legitimate work to proceed with fewer interruptions while enabling review of behavior spread across many requests and sessions. Anthropic requires 30 days of data retention for LSVP traffic. The company said the retained data is compartmentalized, cannot be used for model training, and cannot be accessed by Anthropic’s life sciences research teams.

The beta is available in Anthropic’s first-party console for API usage and in Claude for Enterprise and Team plans. Individual plans, third-party platforms, and BAA-enabled organizations are not yet supported, so customers using PHI data should use separate non-BAA organizations with non-HIPAA data. Anthropic expects to enroll hundreds of organizations within the first week and expand support for most of the life science community in the coming weeks.