AI for Emotional Support?

AI Companions and the Future of Independent Living: Promise, Pitfalls, and Practical Support

Opinion piece contributed by Marco Riedesser, founder of A Friend

Amid an ageing population and a healthcare system under increasingly extreme pressure, the question of how we might support the elderly to live independently at home has become more urgent. Furthermore, loneliness affects millions, particularly those in the later stages of life.

In this context, artificial intelligence is beginning to offer new possibilities, especially in the form of AI-powered companions designed to offer both emotional support and practical assistance.

Whilst developing A Friend, I’ve seen firsthand both the promise and complexity of introducing AI into such a deeply human space.
 

Presence and interaction

Loneliness moves beyond just a social issue; it’s linked directly to health. There has been extensive research linking chronic loneliness to cognitive decline, depression, and physical illness. There are traditional solutions which are effective and worthwhile, such as community programmes or home carers. However, these are difficult to make available at scale and at the frequency required.

Hence, there is a gap in which a physical AI companion may be able to offer support. Unlike the technology of days gone by, modern AI systems can engage in conversation, remember user preferences, and adapt to daily routines. For someone living alone and longing for a little daily interaction, they can help create a sense of companionship. While not directly recreating human support, they can meaningfully reduce feelings of isolation.

It doesn’t require a non-stop back and forth or hours of engagement; even minor, sparse interaction can make a difference. Daily check-ins, reminders, or a little light conversation can make a significant difference. There is even potential for them to act as an early warning mechanism by spotting behavioural changes or being programmed to reach out for help if no contact has been made for a set period of time.
 

Supporting independence

Beyond simple companionship, AI can play a practical role in extending people’s ability to live independently. This is where the tech truly aligns with the goals of independent living.

AI systems can:
• Set reminders for medication and adherence tracking
• Schedule appointments and make prompts for daily routines
• Facilitate communication or health alerts to family members or carers

For many, the issues that lead to us not being able to live independently are multiple and combined, rather than due to any single problem. AI can play a role in helping to manage those daily friction points, reducing a reliance on human supervision, but while still providing a safety net.

There is also a growing appetite for means by which families can have reassurance that their relative is OK, without being intrusive, and maintaining their independence. AI-powered alerts or warnings can become highly valuable in this case.
 

An increasing need

One of the most striking things we encountered during our development and research for A Friend was how widespread and often unspoken the need for support really is. Many, across all ages, are reluctant to admit the need for help, particularly when it might be an indicator of a deeper personal or health issue.

AI can offer a level of support with less associated stigma than some others, in a way not wholly intrusive to your day-to-day life. It can be introduced slowly, framed as a helpful tool, rather than a signal of any wrongdoing or decline. This kind of subtlety is important.

There is even a case for using AI as a preventative, proactive measure, rather than a reactive one.
 

The pitfalls

Despite its clear potential and use case, AI for emotional support is not completely without risk, and it’s necessary to approach this tech with a clear picture of its limitations.

First, AI is not intended, nor should it be positioned, as an entire replacement for human care or relationships. Any support it provides should be supplemental, and not overly relied on in a way which can lead to further societal isolation. The objective here is augmentation, not substitution.

Privacy is another concern. Many AI models are built to collect data and conversational exchanges to train and function effectively. Some don’t collect data at all, but those that do need to ensure that data is handled securely, transparently, and with user consent. This becomes even more vital when dealing with sensitive health and behavioural information.
 

A clear boundary

While this is a piece focused on the potential for AI to help the older among us to maintain their independence, there is a separate issue worth addressing. AI use for children should be limited and overseen by adults. It can be helpful for educational research purposes as an example; however, positioning it as an emotional companion for young people is deeply problematic.

The difference here is that children are still developing socially and emotionally. Introducing AI at an early stage as any kind of replacement for human interaction risks the progress of these developmental phases, and may, in time, lead to confusion in distinguishing between authentic and artificial relationships.

Therefore, it is only among adults that the use of AI companionship should be pursued, given that they are able to engage with the technology in a balanced and informed way.
 

Designing for dignity

Ultimately, the theoretical potential of AI to assist independent living will only come to fruition if the design and deployment strategies of manufacturers are on point.

The best systems will be those that prioritise dignity and transparency with regard to data usage, as well as user control. The idea is that they’re built to empower individuals, not impose rigidity or create new kinds of dependency.

Just as important will be a built-in ability for companions to work directly with caregivers, healthcare providers, and family members. This process needs to be seamless; friction at this point will only cause further disruption and stress.
 

Looking ahead

I am in no way suggesting AI companions as some kind of silver bullet. But what they could undeniably be is a powerful new tool for addressing some of the most pressing challenges in independent living. When carefully and considerately designed, and with correct implementation, they can reduce loneliness, support daily lives, and provide reassurance to both users and family alike.

This is an opportunity. Not to replace human care, but to extend it, making independence sustainable, accessible, and more humane.

The tech is in many ways still in the early stages of development; during that process, the focus in this type of utility must remain firmly on the people it is intended to serve. If we get that right, AI can become more than just a technological innovation. It can become a meaningful contributor to better, more independent lives.
–
 

About the Author

Marco Riedesser is the founder of A Friend, an AI hardware companion designed to bring more natural, personal interaction to everyday technology. He’s a serial entrepreneur and product builder with a track record of founding and scaling hardware technology ventures. He is best known as the founder of CONTROLLINO, which created one of the first industrial PLC platforms fully compatible with the Arduino ecosystem and is used by engineers and makers worldwide.

Across his career, Marco has launched multiple companies spanning automation, electronics, and consumer technology, combining deep technical expertise with a strong focus on practical, real-world applications.
 

Article ends – click to return to top

What do you think about the potential for using AI to support independent living? You can add your comments below

Further reading and resources

AI chatbots “confident but wrong”

MHRA call for evidence on AI

An article from a couple of years ago about using AI in social care

Make sure you don’t miss out on news from Independent Living

return to top