While true it's not a problem if you do it with people who are convinced there's a difference which is the majority in this hobby.
To be clear too I think about 80% of the time I've done blind testing there has been an audible difference (though I recently redid one that I used to get 95% chance I was not gessing and I couldn't tell the difference so there's also a "training" bias).
Also it's common to have you DBT designed so there is a statistically noticable difference in some of the samples being compared to at least illuminate that bias.