Doctors have found a way to identify the initial strain of this lethal, engineered virus. Most mutations have been easily identifiable by an AI-supported system.
Is it plausible that a mutated strain of such a virus could evade detection the first time the system is exposed (i.e. the first time someone infected with this mutation was scanned and tested).
Ideally, I'd like the virus itself to mimic something else in the extracellular fluid or debris, but I am way out of my league in trying to determine what that could be. I've found some ways that viruses disguise themselves from the body's defenses, but not much in the way of evading detection and testing.
If that is completely implausible, I may have to resort to a bug within the AI detection system.
I'm open to other suggestions as well!