AMA has recently clarified the definition: "A problem in the differential diagnosis that represents a condition likely to result in a high risk of morbidity without treatment." If the patient is completely asymptomatic, then you're at minimal risk because they don't have a presenting problem. If the test is being done because the patient has symptoms and thinks they've been exposed, you could call it an undiagnosed new problem--because maybe it's HIV, or maybe it's some other viral infection.
NAMAS also had a great example. Patient has a rash. We'd normally say that's acute uncomplicated. However, it's a circular rash, the patient recently came back from a camping trip, and while showering, they found a tick. Patient needs testing because they could have Lyme disease, ringworm, or dermatitis. It's the context that makes it more complex. So in your example, does the patient have other factors that make the provider suspect HIV or hepatitis?