What makes a political deepfake believable enough to share?

The signal

A newly published study asks a psychologically important question about political deepfakes: what shapes the difference between noticing manipulation, liking what we see and being willing to share it?

Its framing brings three factors into the same analysis — the plausibility of the deepfake, media literacy and the viewer’s prior attitudes — and examines their relationship with detection, liking and intended sharing on social media.

That separation matters. Detecting that content may be synthetic, feeling positively toward it and choosing to circulate it are not the same psychological act.

The psychological reading

Deepfake research can become too narrowly focused on a single test: Can people tell what is fake?

But social behaviour is more complicated than detection accuracy. Sharing can also involve identity, motivation, social signalling and emotional response. A person may doubt a piece of content and still engage with it; another may find it plausible without passing it on.

The useful psychological question is therefore not just whether synthetic media can fool perception, but how perception interacts with prior attitudes and the social motives around circulation.

The evidence boundary

PSYTECH is keeping this brief deliberately conservative.

The current source record establishes the study’s topic and variables, but the publication page available to this newsroom did not provide enough inspectable methodological and results detail for us to state effect sizes, causal direction or population-wide conclusions from the final paper.

An earlier preprint with closely related framing reported a multi-country experiment, but PSYTECH does not assume that every preprint result carried unchanged into the final publication. The final article should govern any stronger claim.

What PSYTECH would watch next

The important details are the sample, experimental design, how plausibility was manipulated, whether sharing was measured as intention or behaviour, and how prior attitudes were operationalised.

More broadly, current research suggests that political deepfakes should not be treated as uniquely persuasive by default; their effects are context-dependent, and identity, trust and motivated reasoning can matter alongside detection.

Sources

- How deepfake plausibility, media literacy, and personal attitudes shape detection, liking, and intentions to share political deepfakes on social media. Humanities and Social Sciences Communications (2026).
- Plohl, N. et al. Earlier preprint: How deepfake quality, media literacy, and personal attitudes shape detection, liking, and social media sharing of political deepfakes (2025).