As AI evolves, it gets better at rendering realistic images. However it seems that there’s a huge gap in who can and can’t identify when something is AI generated, or human made.
Older individuals are often more easily deceived by AI-generated images due to a combination of factors, including limited familiarity with AI technology, cognitive tendencies, and generational differences in digital literacy (Ralls, 2024)
Firstly, many older adults lack exposure to and understanding of AI tools like DALL-E or MidJourney, which produce highly realistic images. Studies show that adults over 50 are less likely to identify such content as computer-generated, often attributing it to human creators. This stems from a significant gap in awareness and experience with AI compared to younger generations who grew up in a digital-first environment (Ralls, 2023). As it is: the battle against false information in the older generation has been ongoing. Plenty of scammers use this to their advantage and target the elderly in their online scams.
Secondly, research suggests that older adults may perceive AI-generated content as more “human-like” than younger individuals do. This could be linked to differences in how they process visual and auditory cues, focusing more on literal details rather than subtle markers like rhythm or intonation in speech or images. This makes noticing things that are “off” about images more difficult. They are less likely to notice an off set pupil, or missing fingers (Jones, 2024). Additionally, older adults exhibit higher false alarm rates when discerning AI-generated visuals, indicating a vulnerability to deception (Velasquez, 2024). This all ties into what we call the “Uncanny Valley Effect”. The underlying affect some feel when looking at humanoid robots. A study found that older generations are less likely to experience this effect, therefore less likely to distinguish real from robot (Tu, et.al, 2020)
Finally, generational differences in media literacy play a critical role. Younger generations are accustomed to questioning online content due to their exposure to misinformation and doctored media from an early age. In contrast, older adults often approach online content with less skepticism, leaving them more susceptible to AI-driven scams or manipulated imagery (Jones, 2024). Addressing this issue requires targeted education initiatives and improved digital literacy programs for older populations. Such examples are suggested below.
Older adults can be better educated about AI-generated content through tailored educational initiatives that address their unique needs and learning styles. Here are some strategies:
- Specialized AI Classes
Community centers, libraries, and organizations like AARP offer free or affordable courses designed for seniors. These classes cover topics such as identifying AI-generated scams, understanding deepfakes, and using AI tools like image generators (Gao et. Al, 2024) - Personalized Learning Tools
AI-powered adaptive learning platforms can tailor content to individual cognitive abilities and interests, making the material more engaging and accessible. These platforms provide immediate feedback and targeted support to enhance comprehension (Spulber et al. 2024) - Intergenerational Learning
Programs that pair older adults with younger generations foster collaborative learning. This approach not only improves seniors’ digital skills but also promotes social cohesion and mutual understanding of technology’s challenges (Spulber et al. 2024) - Practical Skills and Healthy Skepticism
Educators emphasize practical skills like verifying online information and recognizing manipulated media. Promoting a balanced skepticism helps seniors critically evaluate AI-generated content without fear or over-reliance (Beaty, 2024) - Accessible Resources
Free online resources like Senior Planet offer introductory courses on AI, while local institutions provide workshops to ensure seniors have equal access to technology education (Beaty, A) - Social and Emotional Support
Incorporating psychological support into AI education reduces feelings of isolation and builds confidence in using new technologies. This holistic approach ensures older adults feel empowered rather than overwhelmed by AI advancements. (Spulber et al. 2024)
These efforts can collectively enable older adults to navigate AI-driven changes with confidence and critical awareness. Hopefully there is a day where it is more obvious to them what they should -and shouldn’t- believe.
CryptoRank. (n.d.). 2023 The growing concern of AI-generated content and its impact on older users: AI. https://cryptorank.io/news/feed/e2c12-ai-generated-content-impact-on-older-users
Gao, Y., Liang, J., & Xu, Z. (2024, August 16). Digital social media expression and social adaptability of the older adult driven by Artificial Intelligence. Frontiers. https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2024.1424898/full
Jones, N. (2024, September 27). A.I. is heavily affecting our older generations. The Lamron. https://www.thelamron.com/opinion/ai-is-heavily-affecting-our-older-generations
Ralls, E. (n.d.). 2024 Older adults view AI as more human-like than younger adults. Earth.com. https://www.earth.com/news/older-adults-view-ai-as-more-human-like-than-younger-adults/
Beaty, A. 2024 How can seniors learn about Ai’s benefits and threats? take a free class – here’s how. ZDNET. (n.d.). https://www.zdnet.com/article/how-can-seniors-learn-about-ais-benefits-and-threats-take-a-free-class-heres-how/
Spulber, D, Amoretti, G, Siri, A Sciendo. (n.d.). 2024 https://intapi.sciendo.com/pdf/10.2478/gssfj-2024-0001
Tu YC, Chien SE, Yeh SL. Age-Related Differences in the Uncanny Valley Effect. Gerontology. 2020;66(4):382-392. doi: 10.1159/000507812. Epub 2020 Jun 11. PMID: 32526760.
Velazquez E, Flores-Cruz G, Roque N. DETECTION OF AI-GENERATED IMAGES: A MIXED METHODS STUDY ON AGE-RELATED DIFFERENCES. Innov Aging. 2024 Dec 31;8(Suppl 1):1301. doi: 10.1093/geroni/igae098.4158. PMCID: PMC11693250.
Comments
2 responses to “AI and the Elderly: Easily Duped”
Interesting and relevant. On the training aspect, one issue I can see is the adoption rate. Current level of adoption with generative AI at 17% of the population over 50 (Gomes, 2024). Literacy of image creation techniques at lower levels of comprehension, combined with higher levels of visual and cognitive impairment, will somewhat naturally cause resistance in the less literate and older populations.
The good news is that the AARP and OATS organizations have created learning opportunities for Seniors that can help overcome the lack of understanding. This however still requires that Seniors need to be willing to take on learning, which isnt always easy to accomplish (Gomes; 2024).
Gomes, L; (2024); https://www.showmetech.com.br/en/people-over-50-are-more-fooled-by-photos-and-videos-generated-with-AI/
Older people being less able to distinguish between misinformation and truth is limited to the “oldest of old (75+)”. The difference found for all other ages is related to individual differences in analytical ability.
Pehlivanoglu, D., Lighthall, N. R., Lin, T., Chi, K. J., Polk, R., Perez, E., Cahill, B. S., & Ebner, N. C. (2022). Aging in an “infodemic”: The role of analytical reasoning, affect, and news consumption frequency on news veracity detection. Journal of Experimental Psychology: Applied, 28(2), 283–298. https://doi.org/10.1037/xap0000426