Provision record
Anthropic · Anthropic Responsible Scaling Policy · View original document ↗

CBRN Capability Threshold and Safeguard Requirements

High severity High confidence Explicit document language Unique · 0 of 352 platforms
Stay ahead of the changes
Track Anthropic and get the diff the day its terms change.
Share 𝕏 Share in Share 🔒 PDF
Document Record

What it is

The RSP establishes a capability threshold specifically addressing chemical, biological, radiological, and nuclear development capabilities, triggered when a model could substantially uplift CBRN development for moderately resourced state programs, with corresponding required safeguards.

This analysis describes what Anthropic's agreement states, permits, or reserves. It does not constitute a legal determination about enforceability. Regulatory applicability and practical outcomes may vary by jurisdiction, enforcement context, and individual circumstances. Read our methodology

ConductAtlas Analysis

Why it matters (compliance & governance perspective)

This provision establishes that CBRN-related AI capabilities are subject to a defined threshold that triggers mandatory safeguard requirements, reflecting a specific operational constraint on model deployment in relation to dual-use technologies.

Consumer impact (what this means for users)

The RSP establishes that deployment of AI models capable of substantially uplifting CBRN development for state-level programs requires implementation of specific safeguards, which the document states are being designed to prevent misuse in relation to CBRN technologies.

Cross-platform context

See how other platforms handle CBRN Capability Threshold and Safeguard Requirements and similar clauses.

Compare across platforms →
▸ View Original Clause Language DOCUMENT RECORD
"
We have added a new capability threshold related to CBRN development, which defines capabilities that could substantially uplift the development capabilities of moderately resourced state programs. Second, we have disaggregated our existing AI R&D capability thresholds, separating them into two distinct levels (the ability to fully automate entry-level AI research work, and the ability to cause dramatic acceleration in the rate of effective scaling) and have provided additional detail on the corresponding Required Safeguards.

Excerpt from Anthropic's Responsible Scaling Policy

ConductAtlas Analysis

Institutional analysis (regulatory & governance intelligence)

1) REGULATORY LANDSCAPE: The CBRN capability threshold engages with export control regimes including the US Export Administration Regulations (EAR) and the International Traffic in Arms Regulations (ITAR), which restrict the transfer of dual-use technologies.

Insight

Unlock the full institutional analysis

Enforcement risk, jurisdiction flags, contract triggers, and due diligence action items.

Applicable agencies

  • Federal Trade Commission (ftc)
    Oversees unfair or deceptive business practices and can investigate companies that mislead consumers about data collection, sharing, or use.
    Who can file: Anyone affected by the company's practices (US or international)
    What you need: Your account details, a timeline of relevant events, and a description of the specific issue
    What to expect: Complaints inform FTC enforcement priorities and investigations but do not result in individual resolution or compensation
    File a complaint →

Provision details

Document information
Document
Anthropic Responsible Scaling Policy
Entity
Anthropic
Document last updated
May 12, 2026
Tracking information
First tracked
July 12, 2026
Last verified
July 12, 2026
Record ID
CA-P-074230
Document ID
CA-D-00823
Evidence Provenance
Source URL
Wayback Machine
Content hash (SHA-256)
9ad6c66902eb2b771ccc40a9d0e1c2d4664c197372e6833be51b902e16754d55
Analysis generated
July 12, 2026 14:38 UTC
Methodology
Evidence
✓ Snapshot stored   ✓ Hash verified
Citation Record
Entity: Anthropic
Document: Anthropic Responsible Scaling Policy
Record ID: CA-P-074230
Captured: 2026-07-12 14:38:45 UTC
SHA-256: 9ad6c66902eb2b77…
URL: https://conductatlas.com/platform/anthropic/anthropic-responsible-scaling-policy/provision/CA-P-074230/cbrn-capability-threshold-and-safeguard-requirements/
Accessed: Sept. 8, 2026
Permanent archival reference. Stable identifier suitable for legal filings, compliance documentation, and research citation.
Classification
Severity
High
Categories

Other risks in this policy

Get the research letter

Companies change their terms quietly. We read every version and catch what actually changed. One email a week on the changes that matter and what they mean.

Frequently Asked Questions

What does Anthropic's CBRN Capability Threshold and Safeguard Requirements clause do?

This provision establishes that CBRN-related AI capabilities are subject to a defined threshold that triggers mandatory safeguard requirements, reflecting a specific operational constraint on model deployment in relation to dual-use technologies.

How does this clause affect you?

The RSP establishes that deployment of AI models capable of substantially uplifting CBRN development for state-level programs requires implementation of specific safeguards, which the document states are being designed to prevent misuse in relation to CBRN technologies.

Is ConductAtlas affiliated with Anthropic?

No. ConductAtlas is an independent monitoring service. We are not affiliated with, endorsed by, or sponsored by Anthropic.