Claim · 2026-10-28 · snapshot · Stated in source · Supported

Anthropic's transparency hub, as served on 28 October 2026, states that it deployed Claude Sonnet 5.5 with the same chemical and biological misuse protections as Claude Opus 5, treating it as meeting its CB-1 threshold for models able to significantly help someone with a basic technical background make known chemical or biological weapons.

One recorded claim. The claim panel gives its qualification, the source passages with their place in each source, and how it was checked.

What it connects

Transparency hub (Anthropic), as served on 28 October 2026 → states evaluation result → Anthropic's transparency hub states, as served on 28 October 2026, that it treats Claude Sonnet 5.5 as meeting its CB-1 threshold and deployed it with the same chemical and biological misuse protections as Claude Opus 5.

Where it appears

No collection page holds this claim. It is listed with its records in the explorer.

Claim and evidence

Transparency hub (Anthropic), as served on 28 October 2026states evaluation result ↓Anthropic's transparency hub states, as served on 28 October 2026, that it treats Claude Sonnet 5.5 as meeting its CB-1 threshold and deployed it with the same chemical and biological misuse protections as Claude Opus 5.

Machine-checked, not reviewed by an editorStated in sourceSupported

Supporting evidence: primary · 1 source

Date: 2026-10-28 · snapshot
Evaluated system: Claude Sonnet 5.5
What the report measured: Whether the model meets Anthropic's CB-1 chemical and biological threshold
By what kind of method: The page refers to Anthropic's own assessments and to the summaries of key safety evaluations in the model's system card
The result, in the report’s words: Treated as meeting the CB-1 threshold and deployed with the same chemical and biological misuse protections as Claude Opus 5
The caveats the report gives: The page states it gives summaries only and refers readers to the full system card for the complete reported results; the system card itself was not retrieved for this record.

Scope and limitsA summary page, not the system card itself, and a mutable one: the wording may change without notice, and the record is bounded to the snapshot read. It is the developer's own threshold treatment, not an independent finding, and not evidence that the protections work. The threshold is recorded at the generality of the page's own wording.

The evidence 01

primary sourceSupports the claim

Transparency hub (Anthropic), model-by-model safety summaries

Anthropic · Publication date unknown

Cite the original source

Cite the original document, not Selekoa. Check these details against it.

Source excerpt

“On chemical and biological weapons, we treat Claude Sonnet 5.5 as able to significantly help someone with a basic technical background make known chemical or biological weapons, which is our CB-1 threshold.”

Location in source Claude Sonnet 5.5 section, 'Testing Methods and Results' and the chemical and biological summary

How it was read Retrieved directly from the publisher

Machine-checked · 2026-10-30

Source details and check notes

Authors: Anthropic
Version: Retrieved for the candidate pack on 2026-10-28 (SHA-256 ddbdde44443a71db8c87d1d710edb263b5c85583ae4917a9858e5bdbcaaa266e). The page as served on 2026-10-28, a mutable hub that summarises the current model line (Claude Mythos 5.1 down to Claude Sonnet 3.7). It states no publication date for the page itself. Retrieved again at admission on 2026-10-30 (SHA-256 180bb54e99e373f383f2d06ea3975f8819bd2708bc49a9f5fafd088074259a63). Its wording can change; earlier wording is not authenticated.
Canonical address: Open the canonical source ↗
Accessed: 2026-10-28
Language: English (en)
Attribution: Anthropic, “Transparency hub (Anthropic), model-by-model safety summaries”.
Reuse: No open redistribution licence asserted. Bibliographic metadata, editorial paraphrases and short attributed quotations (at most 50 words each). Recheck terms before wider redistribution.
Check notes: Candidate pack passage, AI and biosecurity (research/ai-bio). The quotation (32 words) matched the text retrieved for the candidate pack on 2026-10-28 and again at admission on 2026-10-30, in a fresh retrieval (SHA-256 180bb54e99e373f383f2d06ea3975f8819bd2708bc49a9f5fafd088074259a63; the bytes differ from the pack’s retrieval). Matching ignored differences of whitespace, case, quotation marks, dashes and Unicode compatibility forms, so the source’s typography can differ. The same section states that the model was deployed with the protections used for Claude Opus 5. A cross-family machine check (2026-10-30, item 45) re-read the record’s sources and found full support. No human editorial review.

Status and meaning

Editorial status. Not reviewed by an editor. An exploratory record that no editor has accepted. This status does not say whether a person has read any linked passage. Each evidence record above shows how it was checked.

Relation. A report states the result of an evaluation it ran or commissioned, in its own words, with what was measured, by what kind of method, and the caveats the report itself gives (schema 2.22). The record is the report’s statement, not a verified finding: it says nothing about another system, a later version or real-world risk, and a null or limiting result is not evidence that no effect exists.

Date. The page states no publication or revision date; the record is the snapshot retrieved on 2026-10-28.

Record ID ab-ev-anthropic-hub-sonnet55-cb1