Grief & Loss

AI grief chatbots and “griefbots”: comfort, limits and ethics

AI grief chatbots: what they can offer in mourning, the ethical lines around griefbots that imitate the dead, and how to use them safely.

AI companions in grief come in two very different kinds. The first is the supportive chatbot: a conversational tool that listens, helps you sort thoughts, suggests coping strategies and points you to human help. The second is the so-called griefbot or “deadbot” — an AI trained on a deceased person's messages, voice or photos to imitate them in conversation. The first can be a gentle complement to human support. The second sits on genuinely new ethical ground, and deserves much more caution. This guide looks at both with open eyes: neither hype nor alarm, but the questions worth asking before you lean on either.

What can a supportive grief chatbot actually offer?

Quite a lot, within limits. It offers a place to put words to feelings without worrying about burdening anyone — many grievers censor themselves with family precisely because everyone is grieving. It is available at night, when grief is often loudest. And it can normalise: explaining that waves of grief, anger or relief are common, and that there is no fixed sequence of stages everyone must pass through. The limits are just as real: a chatbot has no clinical judgment, can misread crisis signals, and its empathy is simulated — comforting to receive, but not a relationship. Use it as a supplement to people, never a substitute. If your grief feels overwhelming or you have thoughts of harming yourself, do not rely on a chatbot: call 112 (Sweden, Finland, Denmark, the Netherlands and elsewhere in the EU) or 113 in Norway, or contact your doctor, local health service or a crisis line in your country right away.

What about griefbots that imitate the person who died?

Here honesty requires caution. For some, hearing a simulated voice or reading “new” messages brings comfort and a sense of unfinished conversations completed. But researchers and ethicists raise serious questions: Did the deceased consent to being simulated? Does the simulation delay the natural work of grieving, keeping you in dialogue with a model rather than a memory? Who owns and profits from your loved one's digital likeness — and what happens when the service changes its pricing or shuts down mid-relationship? Researchers at the University of Cambridge have described how such services could cause harm, for example through unwanted notifications, using the simulated person to advertise products, or simulations that are hard to switch off. If you consider one: decide in advance how often and how long you will use it, involve someone you trust, and prefer providers that are transparent about data, consent and exit. And notice how it makes you feel over time — comfort that shades into dependency is a signal to step back.

How do you keep AI in its right place during mourning?

A simple rule of thumb: AI for the practical and informational, humans for the relational, professionals for the clinical. Let technology carry what it carries well — the administration, the questions, the organisation — so your energy stays with the people who share the loss. Choose tools that are transparent about being AI, secure with your data (look for ISO 27001 certification and EU AI Act compliance), and quick to point you to human support. That is the philosophy Solace Care is built on: AI where it lightens the load, humans where it matters, and security you never have to think about.

Frequently asked questions

Are grief chatbots dangerous?

There is not yet clinical evidence that grief chatbots are safe for everyone, and they are not a substitute for professional care — especially if you are very vulnerable. If you use one, treat it as a complement to human support, not a replacement. The risks grow with tools that simulate the deceased, discourage outside help or handle data carelessly. In a crisis, call 112 (113 in Norway) or contact your doctor or local health service.

Is it wrong to find comfort in an AI?

No. Comfort is comfort. The question is balance: AI should widen your support, not narrow it to a screen.

Should I create a griefbot of my loved one?

Move slowly. Consider consent, your own vulnerability and an exit plan first — and ideally discuss it with someone you trust or a grief professional.

Want practical, secure support after a loss — with humans a message away? Create a Solace Care account or read more guides.

Read more

Subscribe image

Be the first to know about latest news & product updates.

Need help with the practical side?

Solace Care shows you step by step what needs to be done after a death, and keeps documents, deadlines and tasks in one place.