With large manufacturing lots, quality control is exercised by random sampling. The frequency of sampling is determined in part by previous history. As the number of incidents of bad parts decreases, the sampling becomes more lax. Modern quality control relies heavily on statistics.
My guess is that the defect arose because of a bad stretch in the stainless steel coil used to make the bolts. Most likely, this was caused by a cold shut in the rod forming process which could go unnoticed by the rod manufacturer. When the rod was converted into bolts, the bad section resulted in a number of bolts in the batch with the defect. These could easily be missed by the quality control sampling.
Bringing this up to the vendor should result in the filing of a non-conformance report to the bolt manufacture who, in theory, will take up correctove actions including increased sampling rates and determining the root cause of the problem.
If it points to the rod manufacturer, they will go through the same process..