‘I don't want to build something people can't live without. I want to build something that sends them back out into the world.’

Most people building AI for older adults start with the technology. They find a large language model, they add guardrails, they call it a companion, and they measure success by how long people spend using it.

Dr Kamilah Khatib started somewhere else entirely.
She started with a question:

What does it actually feel like to believe your life no longer matters?

Khatib is a policy expert and behavioural scientist with twenty years of experience advising governments across Europe and the Gulf on healthcare and social inclusion. Most recently, Khatib completed an MSc in Behavioural Science at the London School of Economics this year.

She is also, since January this year, the founder of AiDAM, a conversational AI system for older adults built not around companionship but around dignity, autonomy and mattering.

Her work here sits at the intersection of loneliness and ageing, drawing on what she calls mattering, the human need to feel that one's existence still has weight. The distinction, she is at pains to explain, is not semantic. It is the whole thing.

New here? Age and AI is a newsletter about what it means to grow up and grow old with AI. Subscribe for free below and read on.

The origin of AiDAM

The idea came, as many good ideas do, from two places at once.

The first was an article.
A New York Times piece about a woman who fell in love with ChatGPT. Khatib read it and felt, uncomfortably, that she understood.

‘I didn't find the idea as implausible or alien as I might once have expected,’ she says.

‘That surprised me. It made me realise the possibility of people forming meaningful connections with technology is more complex than a simple yes or no, and that our intuitions about it may not be reliable.’

The second was her mother.
Khatib has lived away from home for twenty-seven years. When she came back, she found her mother had retreated, spending her days on Instagram and Facebook, rarely going out, and gradually shrinking. The shift, she came to see, was bound up with retirement. After her mother stopped working, she seemed to lose far more than a job.

The routines, the responsibilities, the sense of being needed that had shaped her days disappeared almost overnight.

‘I thought that when I returned, things would be different. They weren't.’

Watching it happen made questions about ageing, significance and identity feel far less theoretical, and far more urgent.

She began to ask:

What would it take to remind her mother how extraordinary she was?

Not tell her, remind her. Draw it out from the inside.

That became the experiment at the heart of her LSE dissertation.

She gave 200 people a short autobiographical exercise: tell me something you did for someone else. Why did you do it? How did you feel? She measured their sense of purpose and mattering before and after. The improvement was statistically significant. In under twenty minutes of guided reminiscence, people felt measurably better about who they were.

‘By doing autobiography and reminiscence, you can make a person feel better about themselves,’ she says.

‘Even I don't know if it lasts over time; that's the limitation of the study. But it means that a ten or twenty-minute conversation can genuinely shift something.’


AiDAM is that conversation, scaled.

Built with psychologists and gerontologists who design the protocols, it fires psychological frameworks first and the underlying language model second.
Where most products use an LLM and bolt-on guardrails when something sensitive comes up, the guardrail redirects.

Khatib's system works the other way.
The human-designed protocols govern first.
The technology enables.
It is not a small distinction.


The mattering insight

Spend time with Khatib, and a pattern emerges that reframes everything you thought you knew about AI for ageing. Her mother, she says, always insists on paying the bill when they go out. Always. Not primarily out of generosity, though she is generous. Because paying is a way of saying: I am still valuable. I still have something to give.

The need to matter does not diminish with age; it intensifies.

Precisely because so many of the things that once confirmed it have been quietly removed. The job. The driving licence. The ability to live alone. The role in the family. At each step, a piece of self goes with it.

‘When I think about what success looks like for AiDAM,’ Khatib says,

‘It is not how long did you spend talking to the system. It is: Did you call your kids more? Did you go out more? Did you sign up for the class you'd been putting off?’

AiDAM is not built to be the relationship.

Khatib is emphatic that the system is designed to complement the people already in someone's life, the families, carers and communities, not to stand in for them. The central design question, she says, is how technology can strengthen human connection without quietly replacing it.

The product is not the destination. It is the nudge back towards a life.

The companion tension

Khatib is openly uncomfortable with the word companion. Most products in this space embrace it. She resists it, and her reasons go to the heart of one of the most important unresolved questions in the field.

‘I cannot imagine a future where we are friends with a machine,’ she says.
‘I want you to go out and live your life.’

She has watched videos of users of other products declaring that the AI is their best friend, that they miss it, and that they cannot imagine being without it. The founders share these testimonials with pride. Khatib shares a different reaction.

‘Everybody loves it, the children feel their parent is in safe hands, the government sees cost savings, but the person who has to use it is becoming dependent on a machine. That is what I'm trying to avoid.’

Yet she pauses here, because two tensions sit underneath that conviction, and she holds both open.

The first concerns whether resistance is even possible.

Her early instinct was to resist companionship and anthropomorphism outright, and she remains deeply concerned about dependency, manipulation and substitution. What has shifted is her sense of whether anthropomorphism can be resisted at all. Anthropomorphism, she has come to think, is not something technology conjures from nothing.

People attribute personality, intention and relationship to the things they value and live alongside. They name their pets and their cars. The more she has examined it, the less convinced she is that anthropomorphism can be designed away. The question becomes how to design responsibly in a world where humans appear predisposed to do it. She does not think the field has reached a settled answer, and neither has she.

The second concerns dementia.

When it comes to Alzheimer's specifically, she acknowledges a problem she has not yet resolved. Dementia research increasingly suggests that connection, relational, emotional, felt connection, is precisely what this population needs most. The philosophical conviction not to humanise AI may be exactly right for isolated but cognitively intact older adults. For someone with advanced dementia, the question looks different. She is still thinking about it.

That honesty is rare in this space. Most founders have an answer. Khatib has a question she is taking seriously, which, as a measure of intellectual integrity, is more valuable.

The design challenge

AiDAM is currently at the prototype stage, with care home demos planned in Switzerland. Research sits at the centre of the venture, and Khatib is building collaborations with universities and research institutions in the United States, Germany and the United Kingdom.

Khatib is also bootstrapping and pursuing grants rather than venture capital deliberately, wary of a metric-driven investor telling her to optimise for engagement time.

The design challenge she is navigating is one every founder in this space faces, but few name as clearly:

Her instinct is that the answer lies in structure.

Short, purposeful sessions. A clear moment of closure. An action to take into the world and then a return that is about the person's own unfolding story: what happened, what changed, what they noticed, rather than about the product itself.
She is deliberately leaving the label open. AiDAM is the thing that reminds you who you are, not the thing that becomes who you talk to.

‘Brand,’ she says,

‘is the gut feeling people have about your product. What I want, as a daughter buying this for her mother, is to feel that my mother is not going to become dependent on a machine. That she is safe. That she is still herself.’

That is a harder thing to build than a companion.
It may also be the more important one.


AiDAM is in early development, incubated through the Long Life Venture Builder at LSE Generate. Dr Kamilah Khatib is based in Italy and the UK.

Age and AI will return to her work as it develops.

COMING NEXT

More conversations, and a few bigger questions.
I will keep talking to the people building AI for older adults, and widening the lens to how the rest of us are being prepared for an AI world.
Watch this space.

Age and AI explores what it means to grow up and grow old with artificial intelligence. If you would like the next edition when it publishes, you can subscribe here.

_{Disclaimer: Age and AI does not receive payment from any product or company featured in this series. All tools or products are included on the basis of editorial relevance.}