Thank you for your interest, and sorry I am replying so late^^
Any toxic effect is always a question of dosage. If you go high enough with the dosage, even the most common compounds will get toxic. If there is data on exposure available, researchers in the lab can either give realistic doses to cells/animals - that means estimated maximum concentrations you would expect to occur. Studies conducted in that way did not show any toxicity of glyphosate.
Or researchers go way higher in concentrations - studies doing this have found toxicity. Also these studies can be valuable, especially when they establish something we call the "no observed adverse effect level" (NOAEL), which is the highest concentration where you can't see a toxic effect.
Agencies like EFSA (europe) or FDA (US) take into account both types of studies, but also epidemiologic studies and exposition data and then try to conclude whether there is a toxicity, and if yes, if it's enough to establish legal tresholds or if there has to be a ban. The latter is mostly the case for carcinogenic compounds.
I will for sure include some more stuff in future posts.
RE: tox-blog #2: Glyphosate and the problem of conflicting studies.