AMR occurs when mutating microorganisms render standard treatments ineffective, and it remains one of the biggest global public health challenges. Progress against it has been slow, held back by fragmented surveillance systems, siloed research efforts, and limited access to integrated, real-world data across healthcare, laboratory, and community settings.

Speaking at the One Health Summit in Lyon, held under France's G7 presidency, Professor the Lord Darzi of Denham, Executive Chair of the Fleming Initiative, described the scale of the challenge as a vast, interconnected web of data made up of millions or even billions of individual signals.

Using AWS's generative AI and cloud services, the Fleming Initiative is bringing together some of the world's most fragmented AMR datasets, including compound libraries and surveillance signals, onto a single cloud-based platform for the first time. The goal is to surface patterns that have stayed invisible while scattered across institutional silos and national borders.

Professor Alison Holmes, Director of the Fleming Initiative, notes that AMR is a global challenge no single institution, country, or dataset can solve on its own, and that the AWS partnership could help connect expertise, data, and technology in previously unavailable ways. The resulting infrastructure is designed to function as a living intelligence platform that grows stronger as more institutions contribute data, supporting researchers, healthcare organizations, industry, and policymakers collaborating across borders, while also building local and regional capacity for AMR surveillance and response.

Key Takeaways

• →The Fleming Initiative is receiving several million pounds worth of AWS cloud and generative AI support to build a global AMR intelligence platform.
• →AMR is predicted to cause 39 million deaths between 2025 and 2050, driven by fragmented surveillance and siloed research.
• →The platform will unify fragmented global datasets, including compound libraries and surveillance signals, for the first time on a single cloud environment.
• →Generative AI is screening a library of over 100,000 compounds, compressing years of lab work into weeks in the search for new treatments.
• →AWS is training foundation models on genomic and surveillance data to build an early warning system for emerging resistance patterns.
• →The initiative aims to boost data collection and computational access in underserved regions disproportionately affected by AMR, while respecting data sovereignty rules.