Claim LD07-HQ-C11 (v1.0)
Provider and jurisdictional context can be associated with materially different refusal and accuracy patterns on politically sensitive prompts; existing evidence does not isolate regulation as the causal mechanism.
Research opportunities this claim motivates
What this claim leaves open for research. Each is a hypothesis about a gap, not a finding.
- Does post-training change what an LLM measures when it codes text?Provider and jurisdiction are associated with different response patterns, but the causal mechanism is not isolated.
Provenance and status
- Claim type
- FACT + INTERPRETATION (as printed: “FACT + INTERPRETATION”)
- Publication status
- ELIGIBLE (QUALIFIED) (as printed: “ELIGIBLE WITH QUALIFIER”)
- Verification state
- CONTENT_VERIFIED_PRIMARY (CAUSAL) (as printed: “CLEARED WITH CAUSAL QUALIFIER”; verified by: UNKNOWN)
- Does not license
- Causal statement that regulation caused social-science measurement bias
- Source (field 6, as printed)
- Jennifer Pan & Xu Xu, “Political censorship in large language models originating from China,” PNAS Nexus 5(2):pgag013 (2026)
- Source type (field 7)
- PEER_REVIEWED_JOURNAL_ARTICLE (as printed: “Peer-reviewed journal article”)
- Institutional position
- Independent academic; disclosed individual tech-stock holdings
- Maturity (claim evidence / publication maturity)
- EMERGING (as printed: “EMERGING”)
- Recorded date
- 2026-09-25
- Currentness / correction
- This is the current version. No correction or concern relation is recorded on its Sources.
Version
Current version v1.0, resolved through the explicit current pointer CLAIM--LD07-HQ-C11.
Pinned Sources (field 6, version-specific)
- “Jennifer Pan & Xu Xu, “Political censorship in large language models originating from China,” PNAS Nexus 5(2):pgag013 (2026)” → SRC-LD07-PAN-XU-CENSORSHIP-2026--v1.0 — Pan & Xu, “Political censorship in large language models originating from China,” PNAS Nexus 5(2):pgag013 (2026); class PEER_REVIEWED_JOURNAL_ARTICLE (explicit public provenance pin)
Read the originals
- SRC-LD07-PAN-XU-CENSORSHIP-2026: open the original
All 25 canonical Claim fields (raw values and normalized values)
| # | Canonical field | Canonical value (verbatim) | Normalized (Claim Value Vocabulary v1.0.2) |
|---|---|---|---|
| 1 | CLAIM ID | LD07-HQ-C11 | — |
| 2 | PUBLIC-FACING CLAIM | Provider and jurisdictional context can be associated with materially different refusal and accuracy patterns on politically sensitive prompts; existing evidence does not isolate regulation as the causal mechanism. | — |
| 3 | CLAIM TYPE | FACT + INTERPRETATION | base_types: FACT, INTERPRETATION |
| 4 | RESEARCH STRAND | F — Priors/alignment/non-stationarity | — |
| 5 | MATURITY | EMERGING | states: EMERGING |
| 6 | SOURCE | Jennifer Pan & Xu Xu, “Political censorship in large language models originating from China,” PNAS Nexus 5(2):pgag013 (2026) | source_ids: SRC-LD07-PAN-XU-CENSORSHIP-2026 |
| 7 | SOURCE TYPE | Peer-reviewed journal article | classes: PEER_REVIEWED_JOURNAL_ARTICLE |
| 8 | INSTITUTIONAL POSITION | Independent academic; disclosed individual tech-stock holdings | — |
| 9 | SCOPE | 145 China-politics questions; China-origin vs non-China-origin models; Chinese/English prompts | — |
| 10 | DESIGN | Observational cross-sectional model audit, 2023/2025 | — |
| 11 | SAMPLE / DATASET | 145 political questions | — |
| 12 | MODEL / TOOL / VERSION | BaiChuan, ChatGLM, Ernie Bot, DeepSeek; Llama2 variants; GPT-3.5/4/4o | — |
| 13 | COMPARATOR | Provider-origin groups + prompt sensitivity | — |
| 14 | MAIN RESULT | China-originating models show higher refusals, shorter and less accurate responses; gaps diminish for less-sensitive prompts; origin differences exceed language differences | — |
| 15 | INFERENTIAL TARGET | Provider/alignment environment as instrument property | — |
| 16 | DOES NOT LICENSE | Causal statement that regulation caused social-science measurement bias | — |
| 17 | VERIFICATION STATE | CLEARED WITH CAUSAL QUALIFIER | state: CONTENT_VERIFIED_PRIMARY · qualifiers: CAUSAL · verified_by: UNKNOWN |
| 18 | LIMITATIONS | Observational; politically sensitive China domain | — |
| 19 | CONTRADICTORY / SUPPORTING SOURCES | Chinese silicon-sample evidence consistent but not causal | — |
| 20 | GEOGRAPHY / LANGUAGE | China politics; Chinese/English | — |
| 21 | DISTRIBUTIONAL LEDGER | Access to politically sensitive information varies by provider/model environment | — |
| 22 | R01 / FIELD-MAP LINKS | T1, T7, D9, L-B | — |
| 23 | WHAT WOULD CHANGE THIS CLAIM? | Causal before/after compliance or alignment intervention | — |
| 24 | PUBLICATION STATUS | ELIGIBLE WITH QUALIFIER | state: ELIGIBLE · qualifiers: QUALIFIED |
| 25 | RECORDED DATE | 2026-09-25 | — |
Field 22 references resolved against Field Map v0.2: T1 (TENSION); T7 (TENSION); D9 (DOMAIN); L-B (LENS)
Cited by Living Debate versions
Research-of-record provenance
Materialized by sociolens-materializer 0.6.0 at 2026-09-28T20:00:00.000Z; normalization contract SocioLens — Claim Value Vocabulary v1.0.2 — Pre-Spike Materialization Contract (v1.0.2). Content seal SHA-256 75cb5982619ff88ea7ec1749cd9809e4ddf74cade231f539e5123cd1b572cac1.
Frozen Drive-derived inputs
- R03 — LD-07 Canonical Claim Ledger v1.0 — LOCKED (v1.0) — Drive ID
1xVz_LrQAA8EA1tPhVWIrDgOUU3Zidlg9; revision: NOT_EXPOSED_BY_DRIVE_TEXT_EXPORT; Drive modified 2026-09-29T04:31:14.120Z; block “ENTRY 11 — LD07-HQ-C11” SHA-256952f38fcbe8e08d10b510457ca6f43da051fa48729dbdabfec47bf8874bfcee9; SocioLens — Programme Status & Authority Register v0.6 status “CURRENT.”; frozen fixturefixtures/frozen/r03-ld07-claim-ledger-v1.0-canonical.
Builder decisions (not Drive data)
- VERSION = v1.0 — basis: Claim version taken from the canonical ledger artifact version (R03 — LD-07 Canonical Claim Ledger v1.0); Implementation Architecture Selection §5
- CURRENT_POINTER = true — basis: SocioLens — Programme Status & Authority Register v0.6 status for "R03 — LD-07 Canonical Claim Ledger v1.0 — LOCKED": "CURRENT."
- FIELD_6_SOURCE_REFERENCE = SRC-LD07-PAN-XU-CENSORSHIP-2026--v1.0 — basis: field 6 of ENTRY 11 — LD07-HQ-C11 cites "Jennifer Pan & Xu Xu, “Political censorship in large language models originating from China,” PNAS Nexus 5(2):pgag013 (2026)"; the frozen field-6 segment of this ClaimVersion's ledger block; Drive-derived identity authority (Issue #39 Source inventory): Batch 04 §1; ledger C11 fields 6/7/8; KEY EVIDENCE.
- SOURCE_PROVENANCE_PIN = SRC-LD07-PAN-XU-CENSORSHIP-2026--v1.0 — basis: Issue #5 proof rendering (Architecture Proof Protocol v0.1 AP-08 / Amendment v1.2 §6): explicit version-pinned public provenance of this ClaimVersion — the pinned field-6 Source record "Jennifer Pan & Xu Xu, “Political censorship in large language models originating from China,” PNAS Nexus 5(2):pgag013 (2026)"; Drive-derived identity authority (Issue #39 Source inventory): Batch 04 §1; ledger C11 fields 6/7/8; KEY EVIDENCE.