A Zimbabwe-developed chatbot is offering pregnancy information and referrals to licensed health workers through WhatsApp, using a database of medically reviewed material rather than general internet content. RFI reports that DawaMom has supported more than 20,000 users, while its account of one urgent case shows both the potential and the limits of AI-assisted care.
Watch Desk analysis
What happened
DawaHealth, a Zambia-based startup, developed DawaMom in 2024. RFI reports that the chatbot is free to access, with paid subscriptions for extras such as personalised monitoring and home visits. It answers questions using retrieval-augmented generation, drawing on a database compiled from medical sources including health ministry guidance, World Health Organization recommendations, medical-school curricula and hospital protocols.
The company says more than 25 health workers translated and validated the material into Shona, Ndebele and Tonga. It also worked with more than 30 women on an English-Shona code-switching dataset, and says the chatbot can respond by voice in a Shona accent. DawaHealth says it has supported more than 20,000 users and is targeting 50,000 by the end of 2026.
RFI describes one user who consulted the service after noticing no foetal movement and was advised to seek medical care immediately. She reached a clinic, but the foetus had died. Her account illustrates why timely advice can matter, but it does not establish that the chatbot caused the outcome or can reliably identify every emergency.
Why it matters
This is a concrete use of AI in a setting where clinic queues and shortages of health workers can make reliable information difficult to access. WhatsApp also gives the service a familiar route to users who may rely on mobile data bundles rather than open web searches.
The important distinction is that DawaMom is described as a way to provide information and help direct people to care, not as a replacement for clinicians. In health, a helpful answer is not the same as a diagnosis, and the boundary matters most when someone needs urgent treatment.
Our read
The design choices are more interesting than the chatbot label: a locally adapted, reviewed knowledge base, several regional languages and a link to licensed health workers. That is a more grounded proposition than asking a general-purpose model to improvise medical guidance from the open web.
But the account is not an independent evaluation of medical accuracy or outcomes. DawaHealth’s user figures and performance claims remain its own, and the reported emergency is a reminder that access to advice cannot guarantee a good outcome. The useful question is whether this kind of tool can help people reach qualified care sooner, without being mistaken for that care.
What to watch
- Whether DawaHealth publishes evidence on accuracy, referrals and outcomes.
- How the service handles emergencies and makes clear when users need in-person care.
- Whether its planned growth to 50,000 users is matched by more clinical support and oversight.
Discussion spark: Should AI pregnancy chatbots be expanded where clinics are stretched, or should they be limited to directing users to clinicians until independent evidence shows they are safe and effective?
Sources and evidence
- Source update (17 October 2026, 13:44 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.