If an action or policy has a suspected risk of causing harm, but the scientific community lacks consensus that the action or policy is not harmful, the burden of proof that it is not harmful falls on those taking the action.