Because humans can't be subjected to experimentation, regulators assessing the safety of a chemical have to weigh whatever other, imperfect, evidence they have at their disposal. Safety assessment is particularly challenging when dealing with a possible carcinogen, such as glyphosate herbicide . It's even trickier, it seems, when the chemical is, like glyphosate, a major money-spinner produced by the powerful biotech industry for use on its commodity GM crops .
There are two important types of evidence available to regulators: laboratory experiments using animal models, and epidemiological studies. Neither can yield definitive conclusions, and their limitations are particularly evident in investigations of suspected carcinogens.
The shortcomings of animal models stem from the biology of cancers.