WIRED

Startup fights AI bioweapons with AI defense

Radical Numerics launches Omnii Defense to detect natural and engineered pathogens using AI, as frontier labs warn their technology could enable dangerous bioweapons.

By ·

Nov 11, 2026

6 cards · 1m 40s · tap to begin

From · · · 1m 40s

Startup fights AI bioweapons with AI defense

Radical Numerics launches Omnii Defense to detect natural and engineered pathogens using AI, as frontier labs warn their technology could enable dangerous bioweapons.

In brief

Radical Numerics launches Omnii Defense to detect natural and engineered pathogens using AI, as frontier labs warn their technology could enable dangerous bioweapons. Radical Numerics announced Omnii Defense, an AI model designed to flag natural and AI-engineered pathogens on a laptop with a USB gene sequencer…

Frontier labs warn of AI bioweapon risk

A biohazard symbol spreading out in replicated pixels but kept in check by other pixels

Frontier AI labs have warned that their technology could help develop dangerous bioweapons. Now a startup is building a defense against that threat.

Radical Numerics launches Omnii Defense

On Wednesday, Radical Numerics announced Omnii Defense, an edge model designed to flag natural and AI-engineered pathogens using a laptop and USB gene sequencer. It runs without internet connection.

Cofounder sees coming biodefense arms race

Eric Nguyen, cofounder and CEO of Radical Numerics, says he sees a coming arms race in biodefense similar to cybersecurity, where threat actors and defenders leverage AI.

“We felt that the biosecurity space was far behind on the defense side.”

How Omnii Defense assigns threat scores

How Omnii Defense assigns threat scores

Omnii Defense makes multiple predictions about a sequence, including odds of it being pathogenic or AI-generated. It assigns a genome a threat score between zero and one combining unnaturalness with functional predictions.

AI models outperform older detection tools

Earlier biodefense tools like Felix compare sequences against libraries of known organisms. AI models learn patterns in sequences associated with biological function, recognizing similarities even when genetic spelling changes.

Detection alone is not enough

Bronwyn MacInnis, director of pathogen surveillance at the Broad Institute of MIT and Harvard, cautions that detecting a signal is still a big leap from determining whether a pathogen has real consequence or was engineered.

What it means

Radical Numerics announced Omnii Defense, an AI model designed to flag natural and AI-engineered pathogens on a laptop with a USB gene sequencer, addressing concerns that frontier AI labs could help develop bioweapons.

Read the original on WIRED

How we make summaries

React

Sign in to react and comment.

Comments (0)

Life is short. Keep it sweet. Respect others' opinions and be kind!

    Recommended next

    More decks on radical numerics and related topics.