My wife is working on her biology PhD, so she listens to a lot of podcasts by researchers and virologists and what not all throughout the day. From what I understand, what's likely at fault here is that people had not made a full recovery from the virus, and the viral response required for a "negative" was set higher than it should have been. User error, basically. It's always possible for a very small percentage of a population to not develop antibodies correctly, but it's far more likely that they just never fully recovered in the first place, and health officials eager to score some wins jumped the gun as soon as they saw the smallest sign of success.