Are you trying to check the spectrophotometer assuming the cuvettes are equal, or the other way round? I presume probably the first.
Is this a double beam spectrophotometer? That works by comparing intensity of light passing through the sample to be measured with another through a reference cuvette. So yes, the effect of covering the instrument seems to indicate stray light. But if, say, they were in balance when uncovered and you seem to have absorption by the sample, you could explain the observations by the stray light entering the pathway of the reference sample more than me one you think you're measuring. Maybe the person who last adjusted the instrument did so without covering it! Anyway a thing surely to do, since you must worry about the cuvettes in themselves, or how they sit exactly in the instrument, is to switch them. If you then have a negative baseline, it's the fault of the cuvettes. If you still have the same positive baseline, it's the fault of the instrument, the electronics of the optics.
If you are able to do anything more come back and tell us any results