How Google DeepMind and Isomorphic Labs are using AI to prevent bioweapons misuse and detect outbreaks faster

Detection: AlphaEvolve and AlphaGenome

• AlphaEvolve is a Gemini-powered evolutionary coding agent that optimizes algorithms for complex scientific problems . In the bioresilience context, it is used to accelerate pathogen surveillance and threat detection by improving computational infrastructure for biological analysis .
• AlphaGenome is a model that deciphers the regulatory code within DNA sequences, predicting gene expression, splicing patterns, and chromatin features . It was published in Nature in January 2026 and is now deployed via API for trusted researchers to detect genetic signatures of emerging biological threats .

Response: IsoDDE drug design engine

Isomorphic Labs' IsoDDE (Isomorphic Labs Drug Design Engine), released in February 2026, is a unified AI system that combines protein structure prediction, ligand binding, and antibody interaction modeling into a single pipeline . IsoDDE roughly doubles AlphaFold 3's accuracy on the hardest ligand-binding cases and is used to rapidly design vaccines and therapeutics in response to outbreaks . The engine is supported by Isomorphic Labs' $2.1 billion Series B funding, which is earmarked for scaling drug discovery .

Fit within Google DeepMind's Frontier Safety Framework

The bioresilience program operationalizes the biological risk domain of DeepMind's Frontier Safety Framework (FSF). The FSF, now at version 3.1 (April 2026), defines CCLs and Tracked Capability Levels (TCLs) for CBRN risks, including harmful manipulation . It is the only frontier safety framework among major AI labs that covers all four suggested risk domains: cyberoffense, CBRN, harmful manipulation, and loss of control . The FSF gates which biological AI models are released and to whom, making the bioresilience program a concrete application of the framework's protocols .

Industry and regulatory context

The announcement follows a call by DeepMind CEO Demis Hassabis for a more "systematic" approach to AI regulation — specifically, a standards body for frontier AI funded by industry and answerable to the U.S. government . The bioresilience program has already built more than 15 partnerships with governments, biosecurity organizations, and researchers . DeepMind's senior director of frontier AI policy noted that frontier AI labs are "in agreement" on the need for rigorous pre-release testing, and expressed hope that the Trump administration's executive order process would produce a "durable framework" for frontier AI testing . Rather than framing biology solely as an AI safety problem, Google is increasingly arguing that frontier AI itself will become a critical biosecurity tool .