Tuesday, October 13, 20264:30 PM - 6:00 PM
Bunche Hall, Rm 10383
About the Talk
A fabrication that convinces nobody has still done damage. Once European publics know that synthetic media exists, denial becomes cheaper for anyone with something to deny, and authentic evidence can be dismissed as AI-generated. Chesney and Citron (2019) named this the liar’s dividend, and Schiff, Schiff and Bueno (2025) have since demonstrated it. We argue that it rests on a wider condition, which we term ambient skepticism, a background doubt about the authenticity of mediated information that attaches to genuine material as readily as to fabricated material. Vaccari and Chadwick (2020) found that political deepfakes generate generalised uncertainty more reliably than false belief. The condition is a property of the information environment rather than a deficit in any individual, which is why media literacy campaigns cannot reach it (Ternovski, Kalla and Aronow, 2022).
Sweden is our test case, and a deliberately unfavourable one. Institutional trust is high, with public service broadcasters trusted by 77 percent of their users (Reuters Institute, 2024); the press system ranks fifth of 180 countries for freedom (Reporters Without Borders, 2026); and a Psychological Defence Agency (MDF, 2026) has operated since January 2022, established in the shadow of Russian information operations against the NATO accession debate (Kalla Fakta, 2024). In January 2026, users of Grok, the AI system built into X, generated sexualised imagery of Deputy Prime Minister Ebba Busch through the platform’s own editing tool (Sveriges Radio, 2026; TV4, 2026). None of Sweden’s defences prevented the harm, or the doubt that followed. If the condition takes hold there, weak institutions cannot be the explanation.
The case also exposes a governance void with implications outside Scandinavia. Swedish criminal law has never been tested against fabricated intimate material, despite a European obligation to criminalise it by 14 June 2027 (Directive (EU) 2024/1385, Art. 5). The European Commission opened proceedings under the Digital Services Act on 26 January 2026, and the European Parliament (2026) approved its position on amendments to the EU AI Act on 26 March 2026. French authorities searched the platform’s Paris offices in the same period. Yet the first binding constraint on the system came on 26 March 2026 from a Dutch civil court, acting on a suit brought by two victim-support organisations and attaching a penalty of €100,000 per day (Amsterdam District Court, 2026). Regulation (EU) 2024/1689 aimed at disclosure arrives after circulation by design (Łabuz, 2025).
Drawing on thirteen interviews with Swedish journalists and political communicators (Njekwa Ryberg, 2026), we show that verification is reactive and individualised while the production of doubt is structural and automated. One participant, a defence researcher, observed that what is tested in Ukraine is already being used against Europe. Professional communicators absorb the epistemic cost on behalf of publics who never see the transaction, which makes the sexualised targeting of women in public life an attack on democratic infrastructure rather than on individual reputations.
About the Speakers
Elizabeth Ashley Fox-Jensen, PhD is Visiting Assistant Professor in Information Studies at UCLA and Principal Investigator of the Deepfake Democracy Project, a Vertically Integrated Project previously hosted at Malmö University’s Sustainable Digitalisation Research Centre. She holds a PhD in Digital Humanities from the University of London and is a Visiting Scholar at NYU Tandon School of Engineering. Her appointment at UCLA is supported by a STINT Teaching Sabbatical Fellowship, and she leads a graduate course on authenticity, AI and the design of digital information environments.
Sue-Ellen Njekwa Ryberg, MSc, MA is researcher in communication and policy at Uppsala University’s Department of Government, with an MA in Communication for Development from Malmö University and an MSc in Global Studies from the University of Gothenburg. A former trainee at the European Parliament working on human rights, migration, privacy and democracy policy, she conducted and analysed the interview corpus presented here.
This talk draws on two forthcoming articles: “Weaponized Uncertainty” (*Journal of Digital Social Research*) and “Believe Nothing” (*Oxford Intersections*, Oxford University Press).
Sponsor(s): Center for European and Russian Studies