Does AI Reduce the Mental Load, or Just Relocate It?
Key Takeaways
- A Time investigation into "Care AI" is right to be sceptical: giving one parent a smarter personal assistant doesn't redistribute the mental load, it just raises her processing speed, the same trap historian Ruth Schwartz Cowan documented with washing machines.
- Most "Care AI" tools, including task-delegation assistants, are built to serve one user faster. That's structurally different from a tool that makes plans and reminders visible to a partner, grandparent or childminder without her relaying them by hand.
- University of Bath and Melbourne research found mothers handle 71% of mentally effortful household tasks and 79% of daily ones, and that even earning six figures doesn't shift that share, only the physical chores.
- No mental load app comparison online currently accounts for grandparents or childminders, despite roughly a quarter of under-5s in England receiving informal grandparent childcare. That's the part of the picture "share it with your partner" advice keeps missing.
You know the feeling of typing "swim kit, packed lunch, permission slip" into your phone at 11pm because if you don't capture it now, it'll be gone by morning. A smarter version of that phone, one that remembers for you, drafts the reply, flags the deadline before you ask, sounds like relief. Time magazine ran a long piece this week asking whether that promise, "Care AI," can actually free women from the mental load of caregiving.
Its answer, in short, is no, not on its own. The article traces the same pattern back to historian Ruth Schwartz Cowan's "More Work for Mother," which found washing machines and vacuum cleaners didn't shrink housework, they raised the standard of what "clean" meant and left the same person responsible for it. Time's conclusion is that redistributing the load needs cultural change, not smarter tools.
That scepticism is correct. But it stops one question short: not all "Care AI" is the same kind of tool doing the same kind of job. Some of it makes one person faster at carrying a load alone. Some of it makes the load visible to more than one person. Those are structurally different mechanics, and mixing them together under one label is exactly why so much "AI for mums" content talks past the real question.
Is the problem AI, or who the AI is built to serve?
It's who the AI is built to serve. A personal AI assistant, the kind you talk to like a chief-of-staff, is built to serve one user: it makes her faster at noticing, planning and remembering, which is genuinely useful, but it doesn't change who has to do the noticing in the first place.
Melo, marketed as MentalLoad.ai, is a clear, honest example of this pattern. It positions itself as an AI you talk to so you can "stop being the human inbox... for everyone in your life." That's a real and understandable pitch. It's also, functionally, a faster version of the same single point of failure: one person's processing speed goes up, but she's still the only node the whole system runs through.
Compare that to a tool built to serve the household, not the individual: a reminder that lands on a partner's phone, a grandparent's phone and a childminder's phone at the same time, without her having to forward it to each of them. That second kind of tool doesn't make the person with the load faster. It changes how many people the information actually reaches. That's not a smarter version of the first tool. It's a different mechanic entirely.
Why do "smart" tools keep failing to reduce the mental load?
Because most of them are built to help the person who's already doing the thinking do it faster, not to move the thinking itself. MIT Technology Review's reporting on chore apps found the same pattern years before "Care AI" existed as a phrase: 86% of Cozi users are women, and researcher Allison Daminger, after years studying household task management, put it bluntly: "I can't think of a time [in my research] where a man made a list for his wife."
That line matters because it names the quiet assumption most family apps are built on. The app assumes a manager and a helper. A helper who completes a task when asked is still outside the mental load for that task; the doing happened, the thinking behind it didn't move. The same reporting found something worth sitting with: some couples abandon task-tracking apps entirely once the app makes the contribution gap too visible to ignore, rather than using that visibility to fix it.
This is what Cowan documented decades before either app existed. A washing machine didn't reduce the work of keeping a household clean, it raised what "clean" meant and left the same person accountable for hitting the new bar, just with a faster tool to hit it with. A faster AI assistant for one parent risks doing exactly the same thing: more gets planned, more gets anticipated, and the standard for "on top of things" quietly rises to match, while the responsibility never leaves the one person holding the phone.
What would actually redistribute the mental load, not just speed it up?
Two things, together, not either on its own: someone else genuinely taking ownership of specific areas, and a shared system that holds the information for whoever needs it, so ownership can actually stick.
The academic evidence on why this is hard is unambiguous. Research from the University of Bath and University of Melbourne, published in the Journal of Marriage and Family, found mothers handle 71% of household tasks that require mental effort and 79% of daily tasks like cleaning and childcare, more than twice the share fathers carry. The researchers call this "gendered cognitive stickiness": even mothers earning six figures see no reduction in their share of the mental load, despite doing measurably less physical housework. Employment changes who does the chores. It doesn't change who does the thinking.
That's the part a faster assistant can't touch. Ownership has to be handed over on purpose, area by area, not inferred from a shared calendar. "The kids' clubs are yours, the booking, the kit, the WhatsApp group, all of it" moves the load. "Can you check the calendar" doesn't, because checking still requires someone else to have already done the noticing and entering.
How does sharing with the wider family network change the outcome?
It changes the outcome because for most UK families, "share it with your partner" was never the whole problem. A quarter of children aged 0 to 4 in England received informal childcare from grandparents in 2024, making grandparents the single most common informal childcare provider, and a Mumsnet survey of parents found more than half of those using any childcare rely on grandparents at some point. Childminders, co-parents across two households, and other regular carers add to that picture further still.
Every mental load app comparison currently online, reviewing tools like Cozi, Fair Play trackers and to-do apps side by side, frames "sharing" as something that happens between two adults in one household. None of them mention a grandparent or a childminder once. That's not a small oversight. For a parent who relies on her mother-in-law for Wednesday pickup, or a childminder for the days she works late, a system that only reaches her partner still leaves her as the single point of contact for everyone else, which is most of the actual coordination.
We don't have UK-specific research yet on whether centralising planning with one parent changes how the load ends up distributed. But the same pattern shows up wherever digital household management has been studied, from the US chore-app research to Danish and Norwegian "digital housekeeping" literature: a tool that concentrates information with one person tends to concentrate responsibility with that person too, whatever the tool's marketing promises.
This is a genuinely different problem from the one most mental load content addresses, and it's worth reading alongside how we've written about explaining the mental load to a partner, because the conversation and the system have to work together. A partner who agrees to own the school run in principle still needs somewhere to actually find the pickup time and the entrance, or the questions come straight back to you.
Maxie was built around that second kind of mechanic, not the first. It isn't a personal assistant that thinks for one parent faster; it's a shared place where school information, plans and reminders reach whoever needs them, a partner, a grandparent, a childminder, without you relaying each one by hand. It won't fix a culture that assumes you're the default manager, and we're not going to pretend it does. Maxie just makes it a bit harder for the load to have nowhere to go but back to you. Try it free, no card required.
Related Resources:
Frequently asked questions
Does using a family organiser app actually reduce the mental load?
Only if the information it holds reaches more than one person. An app that only makes one parent faster at planning still leaves her as the only person who has to do the planning. An app that shares reminders and plans with a partner, grandparent or childminder changes who has to notice and remember things, which is what actually reduces the load.
Why doesn't a to-do list app fix the mental load if one partner still has to set it up?
Because someone still has to notice what needs doing, decide it belongs on the list, and write it down. That noticing and deciding is the mental load itself. A to-do list only moves the doing, not the thinking behind it, unless the list is genuinely shared and both people are expected to add to it and check it.
What is 'gendered cognitive stickiness'?
A term from University of Bath and University of Melbourne research describing how the planning and remembering side of family life stays with mothers even when circumstances change. The study found mothers handle 71% of household tasks requiring mental effort and 79% of daily tasks, and that even mothers earning six figures see no reduction in their share, despite doing less physical housework.
Can AI make the mental load worse instead of better?
Yes, if it only makes the person already holding the load faster and more efficient at holding it. Historian Ruth Schwartz Cowan documented this exact pattern with washing machines and vacuum cleaners: household technology raised standards and expectations faster than it reduced the work, so the person responsible ended up busier, not freer.
What's the difference between task delegation and sharing the mental load?
Delegation means someone else does a task you assigned them. Sharing the mental load means someone else notices the task needs doing, plans it and follows through, without being told. A personal AI assistant can help with delegation. It can't share the noticing, because the noticing only happens inside the one person's head that the assistant is built to serve.
Does sharing a calendar with a partner count as reducing the mental load?
It helps, but only partially. A calendar entry answers 'when.' It doesn't answer 'what time does it finish' or 'which entrance,' which is why so many parents still get asked those questions despite everything being technically in the calendar. Real reduction needs the fuller picture to be visible, not just the date.
How do grandparents or childminders fit into reducing the mental load?
For a huge number of UK families they're already doing a large share of the actual childcare, so a mental load solution that only covers 'you and your partner' misses most of the people who need the information. Around a quarter of children aged 0 to 4 in England receive informal childcare from grandparents, making them the most common informal childcare provider.
What actually works to redistribute the mental load in a family?
Two things together: a genuine cultural shift where a partner takes full ownership of specific areas rather than 'helping,' and a shared system that holds the information for everyone involved, not just the parent who set it up. Tools can support the second part. They can't do the first part for you.