AI is increasingly targeting all human interactions. I am an older adult and somewhat savvy in computer technical issues. I do run my own homelab AI server, which is private, and my questions do not go to the cloud. I see AI and robotic devices moving into caretaking space. This is a multibillion-dollar market with our biggest group of people turning 65 daily in this country for several years to come. I have an Alexa assistant, Google assistant, smartphone, and smartwatch in my home or in my pocket. We have adopted these devices into our daily life with few questions.
There are billions to be made for services to our Older Adult community. Many will try to offer solutions. Many will just be a better mousetrap scam.

As we age, we may not keep up with technology and devices. Our first line of home care in our older years may be AI-enabled devices that monitor for the sound of a person falling, a person calling for help, a heartbeat not being normal, or a change in food intake when reviewing daily online food delivery amounts. These are just a few AI devices that collect data into a cloud data store. What happens with that data, and who has access to it? AI will benefit many; the sky is almost the limit of what it can and will do for us. We cannot set it and forget it when AI is monitoring someone 24/7. There are many ethical challenges for Older Adult care using AI.

---

“Designing for Dignity: Ethics of AI Surveillance in Older Adult Care” is a 2025 opinion/position paper by Jeena Joseph arguing that AI-powered smart home technologies in eldercare — such as fall detectors, motion cameras, emotion-recognition systems, and sleep monitors — are increasingly transforming the elderly person’s home from a sanctuary into a surveillance zone. Rather than enhancing quality of life, these systems erode privacy, autonomy, and dignity by harvesting data continuously, making automated judgments about the user’s behaviour, and acting on those judgments with little or no meaningful consent or transparency from the older adult.

The article presents two real-world case studies to ground these concerns: a California pilot where frequent false alarms caused unannounced social-service visits that left an elderly woman feeling surveilled and powerless, and a dementia facility where emotion-recognition tech generated enough false positives that staff began relying on the system’s cues instead of their own interpersonal interactions with patients — undermining relational care. These examples illustrate a recurring pattern: when AI tools override human judgment or fail to account for user comfort and consent, care becomes control.

To address this drift, Joseph proposes a “Dignity-First” framework built on seven pillars: ongoing (not one-time) consent, data minimization and purpose limitation, user-configurable privacy controls (“privacy zones,” pause schedules), transparent feedback loops (summaries of what data was collected and why), dignity-preserving defaults, regulatory audits of eldercare AI systems, and co-design workshops that actively involve older adults in development. The overarching message is that technology should feel like a companion, not a watchdog — and every design decision must be guided by the question: What does it mean to age with dignity in a digitally mediated home?

✅ Pros (4)

1. Clear, grounded framing of a real problem — The article articulates an understated ethical issue in eldercare tech: surveillance disguised as care erodes autonomy and actually degrades quality of life for many elderly users. The problem is increasingly relevant as smart homes become more common in senior living.
2. Concrete real-world case studies — Instead of abstract debate, the author provides two specific examples (California fall-detection false alarms; dementia facility emotion-recognition overreliance) that make the harms tangible and hard to dismiss as theoretical.
3. Actionable framework, not just critique. — The seven-point “Dignity-First” proposals give engineers, product teams, and policymakers actual design requirements to implement (ongoing consent mechanisms) are specific enough to be actionable: data minimization, configurable privacy zones, transparent audit feedback loops.)
4. Strong emphasis on co-design with older adults themselves (“nothing about us without us”) — recognizes demographic diversity within aging populations including literacy, culture, and disability, rather than treating “older adults” as a monolith.

❌ Cons (4)

1. It’s an opinion/position piece, not empirical research — The claims are persuasive but not supported by data, user studies, surveys, or quantitative evidence. No metrics, no sample sizes, no controlled comparisons.
2. No cost/feasibility analysis. – The article advocates for extensive regulatory audits, periodic compliance certifications, and co-design programs but does not address whether small-scale eldercare providers (often cash-strapped) could actually implement all seven pillars or the costs involved in doing so.
3. Underestimates how much dignity depends on safety — For many elderly individuals, especially those with cognitive decline or dementia, continuous monitoring is what lets them age at home safely. The framing sometimes sets up a false choice between “safety” and “dignity,” when some users genuinely prefer surveillance over the risk of an undetected fall or medical event.
4. No discussion of who gets to judge. – The article says policymakers should audit eldercare AI systems for ethical compliance but doesn’t explain who defines what counts as “dignified” and how diverse viewpoints (cultural, religious, individual) fit into one universal framework.

Link to Original Article