A field guide to AI positions

Anthropic

Sources used: First-hand material.

The map position is our interpretation of the public statements below. Read the method.

Position, range and evidence details

Preferred pace: 0; stated catastrophic-risk concern: 51. These are layout units, not percentages or rankings.

Pace range: -35 to 45; concern range: 32 to 86. Ranges show alternative editorial interpretations, not statistical confidence intervals.

Evidence clarity: medium. Basis: primary. Implementation: published-policy-and-announced-commitment.

Open on the map Reading notes · Structured record

What this position covers

Published company safeguards and its announced embedded-evaluator commitment. [1] [2]

Reference this explanation or suggest a correction

Link to this explanation · Suggest a correction · How corrections work

What to keep in mind

Giving outside evaluators access does not commit the company to slow all frontier development. The Responsible Scaling Policy separates company promises from industry recommendations; implementation is not independently audited. [1] [2]

Reference this explanation or suggest a correction

Link to this explanation · Suggest a correction · How corrections work

What they say about the pace of AI

Appendix A promises development delays in specified competitor-related circumstances, not a general training halt. [1]

Find the passage in the source

Appendix A, Anthropic in the lead and Competitors have strong safety measures

Reference this explanation or suggest a correction

Link to this explanation · Suggest a correction · How corrections work

What they say about the pace of AI

The company commits to work with outside evaluators. Wider limits on developing more capable models require coordination. [2]

Find the passage in the source

Three-step plan and Embedded Evaluators

Reference this explanation or suggest a correction

Link to this explanation · Suggest a correction · How corrections work

What they say about catastrophic risk

The policy assesses catastrophic misuse and sabotage that could increase later global-catastrophe risk. [1]

Find the passage in the source

Section 1 and risk-report requirements

Reference this explanation or suggest a correction

Link to this explanation · Suggest a correction · How corrections work

The wider picture

Promises outside scrutiny and conditional safeguards while continuing development. [2] [1]

Reference this explanation or suggest a correction

Link to this explanation · Suggest a correction · How corrections work

Why the position was reassessed

The previous dot treated evaluator access as pace restraint. Reading the company policy separately supports conditional continuation with a wide range. This is an editorial correction, not evidence of a change of mind. [1] [2]

Reference this explanation or suggest a correction

Link to this explanation · Suggest a correction · How corrections work

What can be checked

The policy publishes risk-report and external-review requirements. Announcing a requirement does not establish that it was met in a particular case. [1]

Reference this explanation or suggest a correction

Link to this explanation · Suggest a correction · How corrections work

Share this page

https://theaiatlas.org/groups/anthropic/

Download a share image · Vector image

Image previews are summaries. Keep the page link so readers can check the evidence.

Sources and what we read

  1. 1. Responsible Scaling Policy, version 3.4

    Read introduction, risk-report requirements, external-review provisions and Appendix A. July 8 is the stated effective date. Separates company commitments from industry recommendations; not an implementation audit.

  2. 2. We Must Pace the Frontier

    The essay itself specifies the month. The September 12 publication day is corroborated by the September 14 press roundup. A proposal and a commitment are not independent verification of implementation.

Edition and machine-readable evidence

Content version 0.19.0. Evidence cutoff 2026-09-15; this does not mean every source was read on that day.

Pinned complete dataset · Complete evidence page · Agent consumption guide