Anthropic Responsible Scaling Policy (RSP) v2 — provisions explorer

A procurement officer evaluating a vendor against Anthropic Responsible Scaling Policy (RSP) v2 can use this page to walk the operative articles + paragraphs the instrument touches, anchored to the primary source. Each row groups every coverage cell that names that article so the reader sees WHICH governance topic each clause speaks to. Use it as a checklist scaffold, not as legal advice (see charter §7.4).

§2

1 cell

governs
Foundation Models / GPAI
RSP v2 §2 — ASL framework applies to frontier model releases
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)

§3

1 cell

governs
Catastrophic & Existential Risk
RSP v2 §3 — ASL-3 / ASL-4 capability thresholds explicitly target CBRN uplift + autonomous-replication
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)

§5

1 cell

governs
Transparency Obligations
RSP v2 §5 — public publication of safety determinations + capability eval methodology
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)

Unanchored citations

7 cells

governs
Agentic AI Governance
RSP v2 — ASL thresholds include 'autonomous AI replication' + agentic capability evaluations
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
Compute-Threshold Reporting
RSP v2 capability evaluations triggered by capability rather than pure compute; compute is one signal
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
International Coordination
Seoul Frontier AI Safety Commitments signatory; coordinates with US + UK AISIs on capability evaluation
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
Open-Weight Frontier Release
RSP applies to Anthropic's models which are closed-weight; framework does not address third-party open release
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
Synthetic Content Provenance
Deployment-stage controls would include content provenance where capability tier requires
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
Contract clause language candidates
Candidate language — not legal advice (see /wiki/charter §7.4)
- Vendor shall provide documentation demonstrating compliance with the transparency / disclosure obligations identified at Anthropic Responsible Scaling Policy (RSP) v2, including (as applicable) machine-readable provenance markers, model cards, and end-user disclosure flows.
implicit
AI in Elections
Anthropic's election-integrity acceptable-use policies apply to deployments; not in RSP itself
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
Compute + Model-Weight Export Controls
ASL-3+ tiers include model-weight access controls (recipient-restriction analog)
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)

Unanchored citations

7 cells

governs
Agentic AI Governance
RSP v2 — ASL thresholds include 'autonomous AI replication' + agentic capability evaluations
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
Compute-Threshold Reporting
RSP v2 capability evaluations triggered by capability rather than pure compute; compute is one signal
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
International Coordination
Seoul Frontier AI Safety Commitments signatory; coordinates with US + UK AISIs on capability evaluation
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
Open-Weight Frontier Release
RSP applies to Anthropic's models which are closed-weight; framework does not address third-party open release
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
Synthetic Content Provenance
Deployment-stage controls would include content provenance where capability tier requires
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
Contract clause language candidates
Candidate language — not legal advice (see /wiki/charter §7.4)
- Vendor shall provide documentation demonstrating compliance with the transparency / disclosure obligations identified at Anthropic Responsible Scaling Policy (RSP) v2, including (as applicable) machine-readable provenance markers, model cards, and end-user disclosure flows.
implicit
AI in Elections
Anthropic's election-integrity acceptable-use policies apply to deployments; not in RSP itself
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)
implicit
Compute + Model-Weight Export Controls
ASL-3+ tiers include model-weight access controls (recipient-restriction analog)
Verbatim text:
Verbatim text not yet transcribed — see primary source.
Open primary source: Anthropic Responsible Scaling Policy v2 (Oct 2024)