A free AI drug-discovery platform opens to researchers fighting malaria, TB and neglected diseases

Some of the world's deadliest diseases attract the least drug-discovery money, because the people they kill are mostly poor. Global research and development funding for neglected diseases reached only about 4.17 billion dollars in 2023, roughly 650 million below its 2018 peak. In March 2026, Medicines for Malaria Venture (MMV) and the AI company deepmirror tried to chip away at that imbalance by launching dd4gh, short for Drug Design for Global Health, a free artificial-intelligence platform for drug discovery.

The platform combines predictive and generative AI. Predictive models analyze large datasets to flag which compounds are most likely to become effective drugs, while generative AI proposes new molecular structures to explore. It uses active learning, improving its predictions as researchers feed in fresh experimental results, and lets scientists either build on pre-trained models from curated MMV datasets or train their own. Uploaded data stays private and is not shared between users.

“Global research and development funding for neglected diseases reached only about 4.”

Crucially, dd4gh is free for eligible researchers, with licenses granted to individuals who work on global-health-relevant diseases and conduct non-commercial research. That design directly targets a barrier that has frustrated scientists in lower-income countries. "AI tools for drug discovery are known to researchers in resource-limited settings, but licensing costs can still put them out of reach," said Caroline Maina, a PhD candidate at the University of Cape Town. A launch workshop in Accra focused on supporting African-based researchers working on malaria, TB and neglected tropical diseases. The Gates Foundation provided partial funding.

The honest framing matters: dd4gh is a tool to accelerate the earliest stages of discovery, proposing and prioritizing candidate compounds that still need years of laboratory testing, safety work and clinical trials before any could become a medicine. AI can narrow a vast chemical space down to the most promising leads, but it cannot run the bench experiments or the human studies that ultimately prove a drug safe and effective. It will not, by itself, close a multibillion-dollar funding gap, and access still depends on researchers having the lab capacity to act on the platform's suggestions. But by putting modern AI design tools into the hands of the scientists closest to these diseases, many of them based in the countries that bear the heaviest burden, free of licensing cost, the initiative helps level a playing field that has been tilted against the world's poorest patients for far too long. It is a small but tangible step toward a future where the diseases of poverty draw the scientific firepower they have always deserved.