DNA sequence screening
AssessTechniques
Checking DNA sequences against safety rules or threat databases to flag risky biological constructs.
Why it's here
Placed in Assess: 1 article(s) of evidence from 1 source(s), led by security coverage, with 1 in the last 30 days. Confidence 24%. Low accumulated evidence, so it defaults conservatively pending more signal.
Evidence (1)
- 6Ars Technica AI·9/30/2026securityGoogle develops watermarking for AI-designed proteins
Google has found a way to watermark proteins designed by AI, aiming to help distinguish legitimate research from potentially harmful biological designs. The move addresses a biosecurity gap, since existing screening tools do not reliably identify AI-generated proteins as threats.