Phase I determines safety. Phase II determines efficacy. Phase III is larger scale, randomized, and blind to give the FDA more information on the range of potential adverse effects and the effectiveness of the drug. FDA is supposed to care about both. The acceptable safety of a drug is relative to the effectiveness of it (i.e. a drug may be more dangerous or have more adverse effects, but it’s deemed “worth it” due to the effectiveness of the drug while of course maintaining a minimum safety risk for all drugs). In other words, both safety and effectiveness are both supposed to be taken as equal halves of the whole by the FDA.