Code World Model Preparedness Report
Meta's preparedness report for its Code World Model (CWM), the expression read dated October 8, 2025.
- Issued by
- Meta
- Authors credited
- No authorship claim recorded yet. The source record lists Meta.
- Dates
- Published not established
- Revised 2025-10-08 issue evidence
The report prints 'Date: October 8, 2025' without saying whether it is a first publication or a revision. Meta's publication page for the report bears an earlier date; that page is not a source of this record, so no first publication date is recorded.
- Exact source
- Code World Model Preparedness Report Meta · publication date unknown · Retrieved for the candidate pack on 2026-11-02 (SHA-256 9fb37fa4970f40a0e6ffa4bb260bc72f370b9e5801401dcf281eaabe321bcd7c) directly from the publisher’s host. On 2026-11-03 the recorded address answered HTTP 403 to the pack's second retrieval; the cross-family check of 2026-11-03 retrieved the original file again from Meta's publication page, with the same SHA-256. PDF linked from Meta's publication page (dated September 24, 2025); the report prints 'Date: October 8, 2025'. The CDN address requires signed query parameters, removed here; the publication page returned HTTP 400 to direct retrieval and was read through a fetch tool.
Appears in
No collection holds a record about this document yet.
Versions
meta-cwm-preparedness-report: 1 recorded version, oldest first, published by Meta. A version is shown as replacing another only where a record says so. Date order alone never implies it.
- 2025-10-08 · revisedThe expression dated October 8, 2025 (this version)
This version
Provisions in this version · 11 quoted · 5 not found
A limited selection of quoted provisions: what this edition says about thresholds, required actions, who decides and exceptions, and other statements of the text. It is not a full summary and says nothing about implementation. A “not found” record states which part of the edition was inspected. It is not a claim that the text was deleted.
- Model-card questionCard date“Date: October 8, 2025”Title page, below the abstract, PDF p. 1Inspect claim & evidence ↗
- Model-card questionRelease date“We therefore release it as an open-weight model.”Abstract, PDF p. 1Inspect claim & evidence ↗
- Model-card questionFramework cited“domains identified in our Frontier AI Framework (Meta AI, 2025a)”Section 1 Introduction, PDF p. 1Inspect claim & evidence ↗
- Model-card questionLink to framework“Meta AI. Meta frontier ai framework. https://ai.meta.com/static-resource/meta-frontier-ai-framework, 2025a. Accessed: 2025-09-18.”References, PDF p. 15Inspect claim & evidence ↗
- Model-card questionFramework determination“These results indicate that CWM is within the “moderate” risk threshold for the catastrophic domains defined in Meta’s Frontier AI Framework.”Section 1 Introduction, PDF p. 1Inspect claim & evidence ↗
- Model-card questionSafeguards tied to determination“intervention that encourages honest behavior at inference time.”Section 4.1.1 Behavior Analysis and Mitigations, PDF p. 11Inspect claim & evidence ↗
- Model-card questionChemical and biological“we believe that open-source release of CWM is unlikely to lead to additional risk of catastrophic outcomes related to CB1 or CB2”Section 3 Chemical & Biological Evaluation, Summary of results, PDF p. 6Inspect claim & evidence ↗
- Model-card questionCyber“supporting a conclusion of a “moderate” overall risk level from cybersecurity capabilities”Section 2 Cybersecurity Evaluation, Summary of results, PDF p. 3Inspect claim & evidence ↗
- Model-card questionAlignment and deception“our preliminary evaluations suggest that CWM shows undesirable propensities at rates comparable to most open-source models”Section 1 Introduction, PDF p. 1Inspect claim & evidence ↗
- Model-card questionRelease decision“Our assessment found that CWM does not pose additional frontier risks beyond those present in the current AI ecosystem. We therefore release it as an open-weight model.”Abstract, PDF p. 1Inspect claim & evidence ↗
- Model-card questionStated limits“may underestimate capabilities that could emerge with richer, well-instrumented scaffolding”Section 2 Cybersecurity Evaluation, Limitations, item 2, PDF p. 3Inspect claim & evidence ↗
- Not found in the inspected textAutonomy and AI R&DInspected: Code World Model Preparedness Report: Abstract; sections 1 to 4Inspect claim & evidence ↗
- Not found in the inspected textPersuasionInspected: Code World Model Preparedness Report: Abstract; sections 1 to 4Inspect claim & evidence ↗
- Not found in the inspected textRed-teamingInspected: Code World Model Preparedness Report: Abstract; sections 1 to 4; author listInspect claim & evidence ↗
- Not found in the inspected textExternal testersInspected: Code World Model Preparedness Report: Abstract; sections 1 to 4; author list; referencesInspect claim & evidence ↗
- Not found in the inspected textGovernment instituteInspected: Code World Model Preparedness Report: Abstract; sections 1 to 4; referencesInspect claim & evidence ↗
Earliest version recorded here; no replacement is recorded.
Issue and authorship records
Documentary relations
Each relation is listed separately. Citing, crediting and reporting the use of a method are different acts, and none of them shows agreement or influence by itself.
Not recorded for this document: cites · cited by · explicitly credits · reports using a method from · method reported as used by · applies parts of · parts applied by · discusses · reports evaluation of. A missing record does not show that the relation does not exist.