POLICY · STARTUPS · #3442
Radical Numerics announces Omnii Defense, an edge AI model to flag natural and engineered pathogens
Radical Numerics announced Omnii Defense, an offline-capable edge genomics model that scores sequences for pathogenicity, taxonomic similarity, and likelihood of being AI-engineered; access is currently limited to vetted external testers while the company works with national labs, government agencies, and independent evaluators. The company previously released Omnii, a multimodal genome language model used to design bacteriophages that were deployed under FDA permission in a patient case.
KEY POINTS
- Radical Numerics announced Omnii Defense, an offline-capable edge genomics model that scores sequences for pathogenicity, taxonomic similarity, and likelihood of being AI-engineered; access is currently limited to vetted external testers while the company works with national labs, government agencies, and independent evaluators.
- The company previously released Omnii, a multimodal genome language model used to design bacteriophages that were deployed under FDA permission in a patient case.
- If effective, an offline AI model that detects novel or AI-engineered threats could complement catalog-based tools and change how biodefense surveillance is done, though performance and real-world utility remain unproven.
WHY IT MATTERS
If effective, an offline AI model that detects novel or AI-engineered threats could complement catalog-based tools and change how biodefense surveillance is done, though performance and real-world utility remain unproven.