Ed Yong writing for The Atlantic:

Small groups of researchers can make virtually unilateral decisions about experiments that have potentially global consequences, and that everyone else only learns about after the fact.