AI tools and services have been all the rage on the Internet recently. You can use AI to remove the noise from your audio recordings, create beautiful edits of the images in your phone, write essays, find new recipes, and do a whole lot of other tasks. On the flip side, malicious entities employ AI tools to create fake videos in order to spread disinformation or harass people. In this article, we bring you the best 4 ways to detect AI-generated deepfake videos on the Internet.
Read Also: 45% of serious gamers in India are earning above 6 lakh per annum: Report
In This Article
Pay close attention to the facial movements
The vast majority of deepfake videos involve superimposing someone’s face on another person. This makes your task easier since you only need to focus on the face to catch any inconsistencies or glitches. Poorly produced deepfake videos can be spotted with ease, as the subject’s lip movement does not align with what they are saying. Other videos may require you to focus more carefully to catch any weird twitches, facial movements, incorrect face shadows, and other cues. One big tell of a deepfake video is that the lighting on the subject’s face is inconsistent with the rest of their body.
Look out for weird cuts and choppy audio
If you frequently watch videos on YouTube, you must be familiar with duplicitous advertisements that employ deepfake videos of popular talk shows to sell you a counterfeit product. The scam artists that produce such videos replicate the background of the TV show or podcast on the advertisement and superimpose the face of the host on the person talking into the microphone. The easiest way to detect these AI-generated deepfake videos is by paying attention to the voice of the speaker. If it doesn’t match the speaker’s actual voice, then you are most likely watching a fake video. If the voice is a perfect match but the audio is choppy, it is most likely edited.
Check the source
If you feel what’s going on in the video is out of character for the person, always check the source of the video. Deepfake videos find virality very quickly and it can be tough to find out where they originated from. If the source of the video is a random account with no credentials or proximity to the subject, the video could be a deepfake. Your next action should be to search for the video on the Internet and check if any major news outlet has verified its legitimacy.
Perform a reverse image search
This method may not work all the time, but you should know about it nevertheless. Since most deepfake videos involve superimposing things on existing footage, a simple reverse image search should bring up the source of the original video. Take as many screenshots of the video as you can and perform a reverse image search on Google to find its source. Read this article to learn other methods for reverse-searching a video.
Frequently Asked Questions (FAQs)
How do I find out that a video is deepfake?
Look out for any weird facial movements, fallacies in lip movement, inconsistent lighting, and flawed audio. If the video you are watching shows any of these signs, it is likely that it is a deepfake.
How do I find the source of a video?
Take several screenshots of different frames in the video and then reverse image search them on Google to find the source of the video.
Is it illegal to make deepfake videos?
Making deepfake videos can attract legal action if they are disseminated for the purpose of harassment, inciting violence or hate, defamation, and other illegal activities.
Read Also: Top 3 ways to contact Apple Support to get help for Apple devices
Conclusion
These are the 4 best ways to detect AI-generated deepfake videos on the Internet. If you’re a Twitter user, always look out for Community Notes under videos that seem to be deepfakes. Once you verify that a video is deepfake, it is a good idea to leave a comment letting other users know as well.
