AI Literacy and How Adults Above 40 Identify and Reason About AI-Generated Content: A Qualitative Study

Sammanfattning

Introduction The rapid development of artificial intelligence and generative artificial intelligence has resulted in an increase in AI generated content spread online. This has raised concerns about the public’s ability to distinguish AI generated content from authentic content on social media. Adults over 40 represent a group that is active on social media but under-represented in research regarding AI detection ability, thus making them a relevant group to study. Research Question How do adults above the age of 40 reason about and interpret AI generated content in comparison to human made content? Methodology A qualitative approach was utilized, ten semi structured interviews were conducted in which participants were exposed to thirteen instances of visual media of which seven were AI generated and six were authentic. This was done without informing the participants of the study’s full purpose beforehand. The collected data from the interviews was analysed by utilizing Braun and Clarke’s six step process for thematic analysis. Results Participants had considerable difficulty distinguishing AI generated content with a relatively low identification rate of approximately 16 percent. Participants mainly relied on implausibility as their strategy for identification rather than applying any systematic scrutiny and the participants’ digital experience emerged as a stronger predictor of identification ability rather than age alone. A paradox of AI awareness was identified where a greater awareness of the technology improved detection ability but also risked resulting in an indiscriminate suspicion towards all content not just AI generated. Discussion The study’s findings suggest that awareness of AI generated content alone is insufficient for a solid detection ability. To improve detection ability, the public need to learn not only that AI generated content exists but also how to think critically about it in a more selective and targeted way.