Researchers showed volunteers a batch of photos and told them to pick out the real ones from those generated by AI. But the deepfake faces were more likely to be seen as real. The scientists aren’t exactly sure why but say the AI faces could match the mental templates. The rise of deepfakes is a problem for online trust and data security. People were more likely to trust information from faces they thought were real, the study found. The number of deepfake photos and videos online is doubling every 6 months. To stay safe, people can reverse image search to check whether a photo is genuine and beware of profiles with too little information or too many followers. Watch the video for more details.