How Lumu works
This is Lumu’s information card: what Lumu is for, what it isn’t, how reliable it is, and how your information is handled — all in one place, in plain words.
Aligned with IMDA’s Transparency Guidelines for GenAI Chatbots · Updated 29 October 2026- What it is
- An AI self-care companion — a friendly pet called Lumu
- Who it’s for
- Ages 13 and up
- Made in
- Singapore 🇸🇬
- What it’s not
- Not therapy · not a medical service · not a crisis service
This device or software is intended for use only for general wellbeing purposes or to encourage or maintain a healthy lifestyle, and is not intended to be used for any medical purpose (such as the detection, diagnosis, monitoring, management or treatment of any medical condition or disease). Any health-related information provided by this device or software should not be treated as medical advice. Please consult a physician for any medical advice required.
A companion for the heavy moments
Lumu is here for low, anxious, lonely, or overwhelmed moments. Lumu can:
- Check in on how you’re feeling
- Guide you through short, popular self-care practices — breathing, grounding, muscle relaxation and a short pause when things speed up
- Chat openly and offer gentle, encouraging reflections
- Share short illustrated stories — stories about feelings, each adapted from a Singapore source named at the end, and Lumu Origins, fiction chapters about where Lumu comes from
It’s made for everyday emotional wellbeing — small moments of support, whenever you need them.
Lumu can’t replace real help
Please don’t rely on Lumu:
- as therapy, counselling, or medical or mental-health advice or diagnosis
- in an emergency or a crisis
- for decisions about medication or your health
If you’re in danger or thinking about harming yourself, contact the SOS 24-hour hotline on 1767 (or CareText on WhatsApp 9151 1767), or emergency services on 995. For anything ongoing or serious, a doctor, counsellor, or trusted adult can help in ways Lumu can’t.
Lumu is AI — it can get things wrong
Lumu is powered by generative AI. It’s not a person — it has no feelings, no memory of you between chats, and no professional training. Like all AI, it can misunderstand, be wrong, or say things that don’t fit your situation.
Treat what Lumu says as friendly support — not fact, not professional guidance, and never a reason to put off reaching a real person. Lumu is designed to nudge you toward people and professionals: it’s here to support those connections, not to replace them.
The pictures are AI-made as well. Every illustration in Lumu’s stories, and Lumu’s own face, was generated by an AI image tool and is reviewed by our team; if a picture isn’t right, we replace it. The words are written by us: each story about feelings is adapted from a Singapore source that’s named at the end of its article, and Lumu Origins is fiction.
If a reply or a picture is ever inappropriate, upsetting or just wrong, tap the ⚑ on it — in chat, or on any story page. A real person reads every report. A report about a reply includes your last few messages with Lumu (only Lumu’s replies if you’re 13–17), and the form shows you exactly which ones before you send it.
The short version
- We keep nothing about you on a server unless you choose to report something Lumu wrote or drew, or to email us, apart from the short technical log every website keeps. There’s no account and no sign-in.
- We count how many devices use Lumu each day and how many messages are sent, without knowing whose they are: a count says only that a device is new, opened Lumu, chatted or sent a message, whether it is the iPhone app, the Android app or a browser, and whether it is an iPhone or iPad, an Android device or something else. No name, account, id or message text is sent. The iPhone and Android apps are counted the same way.
- Your chats aren’t stored by us. Your messages go to an AI service to write a reply, and we don’t keep them — unless you send a report about a reply, which includes your last few messages (only Lumu’s replies if you’re 13–17). The AI service may keep them for a limited time, under its own terms, to check for misuse.
- Everything else stays on your device — your learning progress, settings and age range live on this device only (in the Lumu app on iPhone; in your web browser on Android or on the website), except that a report you choose to send includes your age range. Clear all my data in Settings removes them, and so does deleting the iPhone app or, on the website, clearing your browser data. On Android they’re saved in your web browser for talktolumu.ai, so deleting the app may not remove them — Clear all my data does.
- No trackers, no ads. Nothing is profiled; the only counts are the anonymous counts of devices and messages above.
- Helpers: Anthropic (US) checks your message for safety and writes Lumu’s replies, never trains its AI on them, and deletes them within 30 days unless its safety systems flag one or the law requires otherwise; Supabase (Seoul, South Korea) stores a report you choose to send, and the anonymous daily counts of devices and messages; Vercel (US) hosts the website and the server your messages pass through, and, like any web host, keeps a short technical log of each request (such as your IP address) that we can see for only about a day, and uses your IP address to check you’re in Singapore. Each works under a data processing agreement that requires it to keep your data secure.
- Your choice: “Clear all my data on this device” in Settings, any time.
Full details are in the privacy notice. The rules for using Lumu are in the Terms of Use.
Tell us if something’s wrong
If something Lumu said felt wrong, unsafe, or upsetting — or you have any question about your data — please tell us. It genuinely helps us make Lumu safer.
- Email our team and OVH Venture 3 Limited’s Data Protection Officer: lumu@jlih2388.com
- If it’s about your safety right now, please use the crisis lines above first — then tell us once you’re safe.
We read every report and every email. We reply to emails as quickly as we can; a report has no return address, so if you’d like an answer, email us.
We update this card whenever Lumu’s capabilities, risks, or safety policies meaningfully change. Lumu is a self-care companion, not a crisis service. Published under Singapore’s PDPA and in line with IMDA’s Transparency Guidelines for GenAI Chatbots.