In this work, a voice anti-spoofing dataset was built from samples in Latin America Spanish implementing voice conversion and text-to-speech algorithms and later a test was performed on anti-spoofing models trained on samples in English to see their behaviors with other languages than those in which they were trained.