(2023). The power of Alternative Influence Networks (AIN) for spreading Covid-19 problematic information on Facebook during a year of pandemic. https://doi.org/10.1445/106772
Summary
This paper maps the Italian Alternative Influence Network (AIN) that circulated Covid-19 problematic information on Facebook during the first year of the pandemic (March 2020–March 2021). Drawing on hybrid media systems theory and the literature on Intellectual Dark Web (IDW) micro-celebrities, the authors argue that the spread of covid-skeptic narratives cannot be understood as a purely social-media phenomenon: it depends on the remediation of statements originally voiced by experts, journalists, and public figures in legacy media. This remediation mechanism is particularly potent because it grants authoritative voice to misleading claims while evading conventional fact-checking, which struggles when problematic content consists of verbatim — and therefore “accurate” — reporting of false statements made by public figures.
Key Contributions
- Empirical mapping of the Italian AIN during the pandemic, identifying a distinct IDW sub-cluster alongside Politicians, Journalists, Doctors/Scientists/Experts, and Other Public Figures.
- Application of the comments/shares ratio as a diagnostic for distinguishing amplification (share-dominant) from contestation (comment-dominant) dynamics across actor types.
- Theoretical extension of the hybrid media system framework, foregrounding remediation of expert and celebrity statements as a key vector of misinformation circulation.
- Identification of a fact-checking loophole whereby faithful reporting of false statements by elites launders misinformation through legacy media.
Methods
The authors collected 175,228 Facebook posts via CrowdTangle from a pre-identified cluster of Italian covid-skeptic pages and groups. Capitalized words across post fields were extracted; the 2,000 most frequent terms yielded 254 proper nouns, manually coded into five non-exclusive actor categories. After removing figures with formal public/institutional roles, 79 figures were submitted to k-means clustering. Engagement metrics and a comments/shares ratio (range −1 to 1) were computed per cluster. A stratified random sample (N=150) drawn from the top 10% most-engaged posts — stratified by cluster and by three ratio bands — was then subjected to deductive content analysis.
Findings
- Five clusters emerged: Politicians, Other Public Figures, Journalists, Doctors/Scientists/Experts, and IDW Influencers.
- IDW Influencers posted least (1,649) but achieved high mean interactions (1,030.05) and the most share-dominant ratio (−0.28), consistent with hyperpartisan amplification.
- Politicians dominated by volume (9,879 posts) but with lower per-post engagement; Journalists had the highest mean interactions (2,679.76) and the most comment-oriented ratio (+0.16).
- The most engaging posts came predominantly from mainstream-adjacent journalists (Porro, Paragone) and legacy outlets (La Verità, Libero), with AIN micro-celebrities (Montanari, Messora) secondary.
- Posts frequently remediated TV talk-show clips and newspaper interviews to reinforce covid-skeptic, anti-immigration, anti-government, and conspiracy (e.g., microchip/global tyranny) narratives.
- Expert statements were often weaponized to support skepticism rather than to inform; controversial figures (Galli, Capua) drew more comments than shares, suggesting contested reception.
Connections
This paper extends a line of Italian computational work on coordinated and problematic information sharing on Facebook, connecting closely with Giglietto2019-e9be81c1, whose comments/shares ratio metric it adopts, and with Giglietto2022-0e951ac5 and Giglietto2026-632ef967 on coordinated link sharing. Its focus on alternative influencers and the laundering of fringe claims through legacy outlets resonates with broader work on alternative/hyperpartisan ecosystems such as Frischlich2025-vn, Rieder2026-pp, and Marwick2025-ov, as well as with research on remediation and source hacking in the hybrid media system represented by Starbird2025-jj and Donovan2025-ws.