Can you trust your social media feed? UK public concerned about AI content and misinformation

Can you trust your social media feed? UK public concerned about AI content and misinformation

Hoang Nguyen - May 29th, 2024

Artificial intelligence (AI) is rapidly transforming the way we create and consumer online content. From generating realistic images to crafting news stories, AI tools are making their mark on the digital landscape.

A recent YouGov survey conducted among over 2,000 UK adults sheds light on public perception surrounding AI-generated content, particularly regarding trustworthiness and the potential for misinformation.

Learn more about how consumers feel about AI-generated contentGet in touch with a researcher


Concerns about authenticity and misinformation

The survey reveals a high degree of public concern surrounding online content in general, with 81% of respondents expressing worry about its trustworthiness. This concern extends to digitally altered content, with 76% reporting apprehension about photoshopped images and edited videos. Concerns dips slightly for AI-generated content, though almost three-quarters (73%) of UK consumers still express worry about it compared to 8% who are not concerned.

Women were significantly more likely than men to show concern about the prevalence of AI-generated online content (78% vs. 69%) and digitally altered content (80% vs. 72%).


However, the YouGov data suggests some nuances, especially regarding the spread of misinformation. While two-thirds (67%) of consumers show concern for misinformation from AI-generated content, a larger segment (75%) perceive digitally altered content as a strong contributor to misinformation. In both cases, ABC1s were significantly more likely to think that AI-generated content (70% vs. 62% of C2DEs) and digitally-altered content (79% vs. 69% of C2DEs) strongly contributed to the spread of misinformation. This may echo the current reality, as deepfakes and convincingly altered videos have already been used to manipulate public perception.

Labelling AI-generated content

Efforts to combat misinformation often focus on content labelling, that is, tagging AI-generated content as such. Here opinions are divided, with 50% believing labels would be effective in reducing the spread of misinformation from AI-generated online content while 29% find them ineffective. This mirrors the situation with digitally altered content, where 50% see labels as potentially helpful and 29% disagree.

And while labelling may offer some transparency, there’s a big trust issue with labels in and of themselves. When asked about their perceptions of labels on AI generated content on social media, nearly half (48%) of respondents say they would distrust the accuracy of AI-generated content labelling, compared to just a fifth (19%) who would trust it.

The survey results do show a gap in trust when it comes to age. Young adults ages 16-34 are more than twice as likely to trust content labelling when compared to those 55 and over (31% vs. 12%).

YouGov delved deeper beyond just concern, exploring how individuals might react upon encountering social media posts labelled as AI-generated content. A significant portion, 42%, indicated they wouldn’t take any immediate action. This suggests a degree of neutrality surrounding AI-generated content.


However, a sizable segment (27%) would block or unfollow the account, potentially filtering out AI-generated and organic content from a creator altogether. Small figures for engaging with a post (5%), wanting to see more AI content (2%), and sharing the post (2%) indicate a cautious approach.

There were variations in how age demographics might influence people’s reactions to AI-generated social media posts. Younger consumers (16-24) were more likely to say they would engage with the post (i.e., comment, like, etc.) when compared to all respondents (11% vs. 5%). They were also more likely to say they would share these types of posts with others (5% vs. 2%). Conversely, older consumers (55+) were more likely to say they would block/unfollow the account where they saw labels for AI-generated social media posts (33% vs. 27%).

Request the full data or run your own survey on AILearn more

Methodology

YouGov Surveys: Serviced provide quick survey results from nationally representative or targeted audiences in multiple markets. This study was conducted online on 1-2 May 2024 with a nationally representative sample of 2,128 adults in the UK (aged 16+). Data figures have been weighted by age, gender, education, region and social grade to be representative of all adults in the UK. Learn more about YouGov Surveys: Serviced.