Will the lure of deepfakes prove irresistible to democratic governments? What questions should governments ask—and who in government should be asking them—when a deepfake is being considered?