One way you might be able to determine this is by collecting a large enough sample and run it through Benford's Law to see if the number of leading ones and two

One way you might be able to determine this is by collecting a large enough sample and run it through Benford's Law to see if the number of leading ones and twos are much higher than other digits. It's possible they could be inflating and adhering to Benford's explicitly, but if you find the values are nearly even across 1-9 for leading digits, then it's likely they're fake. This is only because the numbers may be varying number of digits and naturally occurring, so it's unlikely you'd get a false positive.