Reimagining HIV prevention with artificial intelligence
Read the article
As the global HIV response enters its fifth decade, the stakes for its future have never been higher. With health-care budgets tightening, donor priorities shifting, and workforce constraints placing strain on overburdened systems, innovation is no longer optional—it is a lifeline.
Progress on global HIV prevention remains dangerously off course with new infections continuing to outpace progress, even with the acceleration of pre-exposure prophylaxis (PrEP) scale-up since 2022. The global HIV response is also facing serious setbacks. HIV prevention programming funded by the US President's Emergency Plan for AIDS Relief has been halted as part of the US Government's cuts to foreign aid; this, combined with reductions from other donors, threatens to decimate hard-won progress at the very moment when acceleration is needed.
Artificial intelligence (AI) provides a unique opportunity to reimagine delivery systems and increase public health impact, if deployed ethically and at equitable scale. Drawing on insights from a March, 2025, consultation convened by Audere Africa and the Desmond Tutu Health Foundation, the resulting insights in this Comment point to a crucial opportunity where AI can help rewire how HIV services are delivered, making them more effective and more efficient.
In such a precarious resource landscape, AI offers clear and expanding value, particularly when embedded in well designed health systems that prioritise equity. When resourced appropriately, these tools can create a people-centred model for HIV prevention—one that enhances patient engagement, improves effective use, and ensures sustainable, accessible care for everyone.
AI extends the reach of overstretched health-care workers by automating routine functions such as triage, scheduling appointments, follow-up reminders, and adherence checks.4 In Kenya and Nigeria, AI powers telemedicine-based PrEP services that eliminate the need for frequent clinic visits. Clients can conduct self-testing, consult with a provider virtually, and pick up medication from conveniently located community pharmacies. Predictive models could further enhance these programmes by helping providers identify individuals at greatest risk of HIV acquisition, forecast supply needs, and target outreach efforts more effectively.
