In a recent interview, Radical Numerics CEO Eric Nguyen describes bio-security as an AI arms race, arguing that defensive capabilities must advance as fast as the offensive use of genome design. His company is building models that reason through biological sequences — what he calls biological chain-of-thought — and combine multiple data modalities, such as genomic, structural, and functional signals. The goal is to stay ahead of adversaries who might weaponize biology, but the same tools are also meant to help scientists design new genomes for legitimate purposes.
The approach treats biological understanding as a reasoning problem rather than a simple pattern-matching task. By forcing models to articulate intermediate steps in a biological context, the company hopes to make predictions more interpretable and actionable. Nguyen suggests that this level of detail is necessary when the stakes are security and public health.
Because the source is a single interview, there is no independent verification of the company's claims or progress. Still, the framing highlights a growing consensus in the field: AI is becoming central to both biosecurity threats and defenses, and the race is defined by speed and adaptability. Whether the technology delivers on its promise remains to be seen, but the direction is clear — AI is now a core tool in the bio-defense arsenal.