Google's DeepMind has published research on watermarking AI-designed protein sequences using their SynthID technology, embedding identifiable markers without compromising protein function. This addresses growing biosecurity concerns that AI-designed proteins could be used to create toxins or alter viral proteins while evading existing DNA detection tools.
Background
AI-designed proteins have shown major breakthroughs in recent years, including enzymes that digest plastics and block venom proteins, but experts have warned these tools could be misused to create harmful biological agents.
- Source
- Ars Technica
- Published
- Sep 30, 2026 at 11:54 PM
- Score
- 7.0 / 10