[Paper Review] Understanding Older Adults' Perceptions and Challenges in Using AI-enabled Everyday Technologies

Esha Shandilya, Mingming Fan|arXiv (Cornell University)|Oct 4, 2022

Technology Use by Older Adults4 citations

TL;DR

This study investigates older adults' perceptions and challenges with AI-enabled everyday technologies through a survey (N=41) and interviews (N=15). Findings reveal that while older adults are eager to learn and use AI, they face barriers due to lack of learning resources, privacy concerns, and fear of losing autonomy. The paper contributes design recommendations to enhance inclusivity, security, and user control in AI products for older users.

ABSTRACT

Artificial intelligence (AI)-enabled everyday technologies could help address age-related challenges like physical impairments and cognitive decline. While recent research studied older adults' experiences with specific AI-enabled products (e.g., conversational agents and assistive robots), it remains unknown how older adults perceive and experience current AI-enabled everyday technologies in general, which could impact their adoption of future AI-enabled products. We conducted a survey study (N=41) and semi-structured interviews (N=15) with older adults to understand their experiences and perceptions of AI. We found that older adults were enthusiastic about learning and using AI-enabled products, but they lacked learning avenues. Additionally, they worried when AI-enabled products outwitted their expectations, intruded on their privacy, or impacted their decision-making skills. Therefore, they held mixed views towards AI-enabled products such as AI, an aid, or an adversary. We conclude with design recommendations that make older adults feel inclusive, secure, and in control of their interactions with AI-enabled products.

Motivation & Objective

• To understand older adults’ personal experiences with AI-enabled everyday technologies.
• To identify challenges older adults face when using AI-enabled products.
• To examine older adults’ general perceptions, knowledge, and (mis)perceptions of AI technology.
• To inform the design of AI products that are more accessible and acceptable to older adults.
• To address the gap in understanding how older adults perceive AI beyond specific products like chatbots or robots.

Proposed method

• Conducted an online survey with 41 older adults (60+ years) who had experience using AI-enabled products to screen and recruit participants.
• Selected 15 survey respondents for in-depth semi-structured interviews to explore their experiences and perceptions in detail.
• Analyzed interview data thematically to identify patterns in perceptions, challenges, and expectations regarding AI.
• Used qualitative analysis to extract design implications based on participants’ feedback on usability, privacy, and control.
• Explored participants’ lack of formal AI knowledge and its impact on their trust and adoption of AI technologies.
• Developed design recommendations based on findings, including printed manuals, progressive disclosure, and privacy transparency.

Experimental results

Research questions

• RQ1 What are older adults’ personal experiences with AI-enabled products?
• RQ2 Do older adults face any challenges in using AI-enabled products? If yes, what are they?
• RQ3 What are older adults’ perceptions of the AI technology in AI-enabled products?

Key findings

• Older adults expressed enthusiasm for learning and using AI-enabled products, but lacked accessible learning avenues or formal education on AI.
• Participants associated AI with 'smartness' in products but often failed to recognize AI when it operated invisibly in background functions like search engines or navigation.
• Many older adults perceived AI as a potential threat to privacy, especially when it collected data, created targeted ads, or enabled sophisticated scams.
• Participants expressed concern that AI could undermine their decision-making autonomy, especially when AI systems acted unpredictably or contradicted their expectations.
• Older adults, particularly those aged 70 and above, preferred printed, well-structured manuals over digital resources, citing difficulty navigating online instructions.
• Participants desired clear, tiered documentation—covering basic, intermediate, and advanced usage—along with guidance for troubleshooting and seeking help.