The Governance-Embedded Interactive Media Effect: the role of AI-generated disclosure in user credibility and engagement based on fact-checking videos on Chinese TikTok (Douyin)

Summary

Note: the structured summary provided describes a different paper than the title suggests — it concerns European fact-checkers’ rhetorical defenses of platform funding rather than AI-disclosure effects on Douyin. Based on the substantive content provided: this paper investigates how European professional fact-checkers rhetorically justify accepting funding and partnership from major platform companies (Meta, Alphabet, ByteDance) while claiming editorial independence. Drawing on 12 semi-structured interviews with fact-checkers from 12 European countries, the authors apply Ware and Linkugel’s apologia theory in combination with journalistic boundary-work theory to identify four modes of differentiation and three modes of transcendence that constitute a predominantly defensive rhetorical repertoire. The paper reframes these defences as boundary work that constructs epistemic authority in a platform-dependent media ecosystem.

Key Contributions

• First application of rhetorical apologia theory to fact-checkers’ defences of their platform partnerships.
• A typology of four differentiation modes and three transcendence modes for analyzing institutional self-defense in journalism-adjacent contexts.
• Extends Graves et al. (2023) by redirecting boundary-work analysis from internal professional boundaries to external boundaries between fact-checkers and platforms.
• Fills a geographic gap by centering European fact-checkers, who remain underrepresented despite Europe hosting the largest number of fact-checking institutions.
• Integrates apologia theory with Karlsson and Örnebring’s “wall” and “membrane” metaphors of journalistic autonomy.

Methods

Twelve qualitative semi-structured interviews (Nov 2023 – Mar 2024) with one professional fact-checker each from twelve EFCSN/IFCN-certified organisations across twelve European countries, all having received platform-company funding between 2021 and early 2024. The interview guide covered six themes (organisational structure, audience, societal role, editorial independence, potentials/challenges, and the role of platforms), averaging 71 minutes. Transcripts (≈107,500 words) were coded in NVivo through four phases: inductive theme coding, theme-based coding, application of the apologia framework, and quote grouping. Interviewees and organisations were anonymised to mitigate harassment and partnership risks.

Findings

• Four modes of differentiation: (D1) fact-checkers vs. platforms to assert independence; (D2) between platforms — often rejecting TikTok/ByteDance while accepting Meta/Alphabet; (D3) between platforms as institutions and well-intentioned individual employees or divisions; (D4) platform funding vs. state funding, defending the former as less compromising.
• Three modes of transcendence: (T1) appeal to a counterfactual worse world without fact-checkers; (T2) appeal to broad anti-disinformation coalitions (“we can’t fight alone”); (T3) appeal to improving platforms from within.
• Funding distribution across the sample: Meta (9), Alphabet (8), ByteDance/TikTok (2), Twitter (2), Spotify/Kinzen (2).
• Fact-checkers report frequent harassment and accusations of censorship, collusion, or serving “globalist” interests, with platform or Soros-linked funding routinely cited as supposed evidence.
• Boundary-drawing is inconsistent across organisations: some refuse TikTok or state money while others accept both, so defences often implicitly critique peers — signaling an emergent, contested field.

Connections

This study connects most directly to work on the political economy and institutional position of fact-checking, including Cazzamatta2026-lo and Dierickx2026-tw on fact-checking practices, and Schiffrin_undated-gi on funding/governance of information intermediaries. Its concern with platform-dependence and editorial autonomy resonates with platform-governance scholarship such as Helmond2026-ll and Rieder2026-pp, Rieder2025-ju, while its attention to the contested legitimacy of counter-misinformation institutions complements Starbird2025-jj, Marwick2025-ov, and Donovan2025-ws on harassment and delegitimisation campaigns against researchers and fact-checkers. It also speaks to debates over the effectiveness and scope of fact-checking interventions explored by Allen2025-ot and DeVerna2025-dl.

Podcast

A research-radio episode discusses this paper: Listen