How Different Age Groups Fare Against AI
As artificial intelligence becomes increasingly blended into everyday life, distinguishing between human-created and AI-generated content is becoming a valuable skill. A new study, Spotting the Bots, explores how different generations perform when faced with this challenge. From social media posts and images to advertisements, participants were asked to identify whether content was produced by a person or an AI system. The results reveal notable differences between age groups, offering insight into how experience, technology exposure, and digital literacy influence our ability to recognize artificial intelligence in the modern world.
1-10: Having only one result in the age 1–10 category, participants performed well at identifying the differences between AI-generated images and text. When presented with human-created content, they were generally confident in recognising it as human-made. This suggests that this age group has a strong awareness of AI and its characteristics. It is likely that this is because they have grown up in a world where AI has become increasingly common and accessible.
10-20: The most exposed generation of people, constant AI in daily life blending with reality. Multiple students anonymously added results to our chart, they show a strong understanding but get weaker on more realistic images made by Sora AI 2. They show a strong understanding of AI images but show a weak trust towards human images and text. Growing up with AI has most likely led to a lack of trust between them. AI text is also a strong point but falls short with any sort of humaniser. These results were most likely caused by the fact they live in the age of AI and constant exposure.
30-50: The results from participants aged 30–50 indicate that this age group is generally effective at identifying AI-generated content. Based on the survey data, respondents achieved an average score of 7.3 out of 10 when identifying AI-generated texts, compared with an average score of 5.1 out of 10 when identifying human-written texts. This suggests that participants found it easier to recognise AI-generated writing than to correctly identify content produced by humans. A possible reason for this is that AI-generated text often follows predictable patterns, uses formal language, and maintains a consistent structure, making it easier to detect. In contrast, human-written responses can vary significantly in style, tone, and quality, which may have made them more difficult to classify accurately. The findings also suggest that adults aged 30–50 possess a good level of digital literacy and are familiar with the increasing presence of AI in everyday life through technology, social media, and workplace tools. However, the lower score for human-written content demonstrates that distinguishing between human and AI-generated text is still challenging. Although the sample size for this age group was relatively small, the results indicate that adults aged 30–50 have a moderate to high ability to identify AI-generated content, while the growing sophistication of AI systems means that accurately distinguishing between human and AI writing is becoming increasingly difficult.
60-80:
In this age group, many people are becoming grandparents and may not naturally recognise AI-generated content. However, because they often spend time with their grandchildren, they are frequently taught about AI and how it works. As a result, many are able to recognise AI-generated images and text more accurately than expected.
80-100: Only one result was reported in this category but, this lady shows the deep trust of the elder community representing the over trust of the recent AI burst. Ann Maison age 98 showed a poor understanding of AI but a very accurate understanding of human images. This may have been because of an overall low average guess and a even lower text guess. Although one result was recorded multiple older people also said and added results to the table, even if it was only one question. All following a similar overly trustful result. This result most likely is happening because of a lack of use of ai and a childhood and adulthood of minimum exposure.
In conclusion, the ability to recognise AI appears to follow a hill-shaped pattern. Young children have little experience with AI, making it difficult for them to identify AI-generated content. As people get older, their experience with technology and AI increases, improving their ability to recognise it. However, this ability appears to decline in older age groups, particularly around the age of 60 and above. This may be because they tend to be more trusting of online content and have had less exposure to AI throughout their lives, as the technology is relatively new. In addition, many older adults use social media less frequently, giving them fewer opportunities to encounter and learn about AI-generated content.
Reported By Sebbi T and Henry D-H