Claim LD10-HQ-C08 (v1.1)

In four preregistered 2025–26 UK experiments, frontier AI out-persuaded strong human text persuaders; its advantage over coached elite debaters disappeared when AI was constrained to the debaters’ message length and writing speed.

Research opportunities this claim motivates

What this claim leaves open for research. Each is a hypothesis about a gap, not a finding.

An updated meta-analysis of AI persuasion that separates attitudes from behaviourNewer UK experiments found frontier AI out-persuaded strong human text persuaders, with the advantage disappearing under human-like length and speed limits; message length and speed are therefore moderators a new synthesis should code.

Provenance and status

Claim type: ESTIMATE (as printed: “ESTIMATE”)
Publication status: ELIGIBLE (PREPRINT) (as printed: “ELIGIBLE WITH PREPRINT LABEL”)
Verification state: CONTENT_VERIFIED_PRIMARY (PREPRINT) (as printed: “CONTENT VERIFIED PREPRINT”; verified by: UNKNOWN)
Does not license: Population-scale influence, face-to-face superiority or election outcomes
Source (field 6, as printed): Hackenburg et al. 2026, arXiv:2606.16475
Source type (field 7): PREPRINT (as printed: “Preprint”)
Institutional position: Academic + UK AISI
Maturity (claim evidence / publication maturity): EMERGING (as printed: “EMERGING”)
Recorded date: 2026-10-13
Currentness / correction: This is the current version. No correction or concern relation is recorded on its Sources.

Version

Current version v1.1, resolved through the explicit current pointer CLAIM--LD10-HQ-C08.

Versions:v1.0 v1.1 · current

What changed in this version

This version supersedes LD10-HQ-C08--v1.0, which stays available unchanged.

Changed fields: 2 PUBLIC-FACING CLAIM; 25 RECORDED DATE
What changed: The public-facing claim now says that the advantage disappeared against coached elite debaters, the only comparison tested under the length and speed constraint, instead of implying it disappeared against all strong human persuaders.
Why: A claim check on 13 October 2026 found that the HQ notes report the constrained result only against coached elite debaters (0.0pp, 95% CI −1.7 to +1.6), while version 1.0 attached it to the advantage over all strong human persuaders. The check and this revision were done by AI agents against the HQ notes, not by a person.

Pinned Sources (field 6, version-specific)

“Hackenburg et al. 2026, arXiv:2606.16475” → SRC-LD10-HACKENBURG-ARXIV-2606-16475--v1.0 — Hackenburg et al. 2026, arXiv:2606.16475; class PREPRINT (explicit public provenance pin)

Read the originals

SRC-LD10-HACKENBURG-ARXIV-2606-16475: 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	LD10-HQ-C08	—
2	PUBLIC-FACING CLAIM	In four preregistered 2025–26 UK experiments, frontier AI out-persuaded strong human text persuaders; its advantage over coached elite debaters disappeared when AI was constrained to the debaters’ message length and writing speed.	—
3	CLAIM TYPE	ESTIMATE	base_types: ESTIMATE
4	RESEARCH STRAND	A — Controlled persuasion	—
5	MATURITY	EMERGING	states: EMERGING
6	SOURCE	Hackenburg et al. 2026, arXiv:2606.16475	source_ids: SRC-LD10-HACKENBURG-ARXIV-2606-16475
7	SOURCE TYPE	Preprint	classes: PREPRINT
8	INSTITUTIONAL POSITION	Academic + UK AISI	—
9	SCOPE	UK Prolific; 10 policy stances; 2025–26	—
10	DESIGN	Four preregistered RCTs	—
11	SAMPLE / DATASET	18,978 conversations; 6,923 people	—
12	MODEL / TOOL / VERSION	Claude Opus 4.1/4.6, ChatGPT-4o, GPT-5.4, Grok 4.20, Gemini 2.5 Pro	—
13	COMPARATOR	Laypeople, selected laypeople, elite debaters, professional canvassers	—
14	MAIN RESULT	AI +8.2pp vs random laypeople; +5.6pp selected; +4.6pp elite debaters; throughput-constrained vs coached debaters 0.0pp [−1.7,+1.6]	—
15	INFERENTIAL TARGET	Newer-vintage AI-human persuasion gap and throughput mechanism	—
16	DOES NOT LICENSE	Population-scale influence, face-to-face superiority or election outcomes	—
17	VERIFICATION STATE	CONTENT VERIFIED PREPRINT	state: CONTENT_VERIFIED_PRIMARY · qualifiers: PREPRINT · verified_by: UNKNOWN
18	LIMITATIONS	Paid opt-in text setting; not peer-reviewed	—
19	CONTRADICTORY / SUPPORTING SOURCES	C07 older meta-analysis shows vintage contrast	—
20	GEOGRAPHY / LANGUAGE	UK / English	—
21	DISTRIBUTIONAL LEDGER	Frontier-model access and throughput may advantage resource-rich communicators	—
22	R01 / FIELD-MAP LINKS	D7, D3, T1, R-1	—
23	WHAT WOULD CHANGE THIS CLAIM?	Peer-review revision or naturalistic replication	—
24	PUBLICATION STATUS	ELIGIBLE WITH PREPRINT LABEL	state: ELIGIBLE · qualifiers: PREPRINT
25	RECORDED DATE	2026-10-13	—

Field 22 references resolved against Field Map v0.2: D7 (DOMAIN); D3 (DOMAIN); T1 (TENSION); R-1 (LENS)

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 2f11b52e4a241d8c8ab8b3cfd315acab3af7fcffbb94e9dff1b1fb2dcd9a7099.

Repository-authored inputs (not Drive exports)

SocioLens claim revision — LD10-HQ-C08 v1.1 (v1.1) — written in the SocioLens repository by Project Lead (Claude, AI) on 2026-10-13; no Drive identity exists. Authority: AGENTS.md "Authority" (the Project Lead decides research content within the research-integrity rules) and docs/PROJECT.md "Research revisions" (a published claim changes only through a new version that states what changed and why). Block “ENTRY 08 — LD10-HQ-C08” SHA-256 86c3b7759cbc4940b366e84e647ff583de78a1fe52330d0f11926f98765d9ade; fixture fixtures/authored/claim-revisions/LD10-HQ-C08--v1.1.

Builder decisions (not Drive data)

VERSION = v1.1 — basis: Claim version taken from the repository-authored revision fixture version (SocioLens claim revision — LD10-HQ-C08 v1.1); a new version of LD10-HQ-C08 v1.0, which stays routable (AGENTS.md research integrity rule 3)
CURRENT_POINTER = true — basis: repository-authored revision "SocioLens claim revision — LD10-HQ-C08 v1.1" (Project Lead (Claude, AI), 2026-10-13); basis: AGENTS.md "Authority" (the Project Lead decides research content within the research-integrity rules) and docs/PROJECT.md "Research revisions" (a published claim changes only through a new version that states what changed and why)
FIELD_6_SOURCE_REFERENCE = SRC-LD10-HACKENBURG-ARXIV-2606-16475--v1.0 — basis: field 6 of ENTRY 08 — LD10-HQ-C08 cites "Hackenburg et al. 2026, arXiv:2606.16475"; the first field-6 segment of this ClaimVersion's repository-authored ledger block; the same Source record LD10-HQ-C08 v1.0 cites.
SOURCE_PROVENANCE_PIN = SRC-LD10-HACKENBURG-ARXIV-2606-16475--v1.0 — basis: explicit version-pinned public provenance of this ClaimVersion (Amendment v1.2 §6): the pinned field-6 Source record "Hackenburg et al. 2026, arXiv:2606.16475".