There's usually an institutional review board that you go through when using research subjects, but you don't need to do that to publish something like a new al

There's usually an institutional review board that you go through when using research subjects, but you don't need to do that to publish something like a new algorithm you developed in a CS journal. In this case, the researcher probably should have gotten IRB approval because there were test subjects--the journal editors, peer reviewers, and probably journal readers. But I'm assuming they just skipped that process and said there were no human subjects.