As has been extensively discussed, mis/disinformation and online harms are an acute threat to our society. This study aimed to build upon the existing body of evidence by answering a direct question: when Australians browse harmful content in a simulated social media feed, what do they actually do?
To answer this question, Verian built a high-fidelity simulated social media platform - closely resembling a real feed - and embedded it within a 20-minute online survey of 1,251 Australian citizens and residents. Rather than asking people how they think they would behave, we observed what they actually did: what they ‘liked’, ‘loved’, ‘shared’, ‘hid’ and ‘reported’. We paired these behavioural measures with demographic measures and established psychometric instruments to identify who is most vulnerable, which content most effectively farms engagement, and what protects against harm.
What we found
Exposure to online harms is near-ubiquitous, and susceptibility is real with a clear demographic and psychographic fingerprint. Within the simulated feed, a majority positively engaged with harmful content - that is, they ‘liked’, ‘loved’ or ‘shared’ posts spanning disinformation, deepfakes, scams, doxxing and ‘manosphere’ material. Perhaps most strikingly, one quarter of participants ‘liked’ or ‘loved’ an explicitly Islamophobic post, with similar proportions engaging with other Islamophobic posts - suggesting that race and/or religion may be a social fault line open to exploitation by bad actors.
Susceptibility varies by age and by a measurable psychological trait: cognitive reflection. Notably, there were few significant differences across geographies - susceptibility is broadly distributed across the country, suggesting a large and pressing national issue rather than a localised one. Participants scoring higher on the extended Cognitive Reflection Test were significantly more likely to take defensive actions (‘blocking’, ‘hiding’, ‘reporting’). Given that literature suggests cognitive reflection can be enhanced via training, this points to a concrete, evidence-based lever for intervention.
Why this matters
The findings of this study establish both a baseline and a method for future work. Beyond the headline prevalence figures, the study demonstrates a replicable, ecologically valid approach to measuring online-harm susceptibility - and to evaluating the interventions designed to reduce it.
Frequently asked questions
What percentage of Australians have experienced online harm? 85% of Australians reported exposure to at least one online harm, such as scams, trolling, hate speech or deepfakes, in the past three months, according to Verian's 2026 research.
Who is most likely to engage with harmful content online? Positive engagement with harmful content rises modestly with age, while a minority of "super-sharers" in every age group account for a disproportionate share of engagement. Lower cognitive reflection is associated with higher harmful engagement and lower protective action.
What is technology-facilitated abuse? Technology-facilitated abuse refers to harms enabled by digital tools, including deepfakes, sextortion, nudify apps and non-consensual intimate imagery. Verian found 86% of Australians are aware of at least one form, but 79% would not know where to seek help if affected.
Do Australians support regulation of online harms? Yes. Verian's research found that over 70% of Australians support regulatory measures including platform fines, bans on nudify apps, and criminalising the creation of non-consensual imagery.
Back
_16x9.png)
