What are the most effective deepfake detection techniques for journalists?

1 Visual clues

By using your own eyes and common sense, or tools like InVID or Forensically, you can detect deepfakes by looking for visual clues that reveal inconsistencies or anomalies in the image or video. These clues may include facial distortions such as blurred edges, mismatched lighting, or unnatural expressions; eye movements such as lack of blinking, irregular pupils, or unrealistic reflections; lip synchronization such as delays, gaps, or misalignments between the audio and the video; and background details like missing or duplicated objects, warped perspectives, or sudden changes.

3 Machine learning models

A more advanced way to detect deepfakes is to use machine learning models that can compare the image or video with a large database of real or fake examples and generate a score or a label that indicates its likelihood of being authentic or manipulated. For example, FaceForensics++ is a framework that can detect and classify deepfake videos based on facial features and movements; DeepFakeLab is a tool that creates and detects deepfake videos using various algorithms and techniques; and Deepware Scanner is an app that scans and verifies images and videos using a neural network and a blockchain system. You can use these models online or offline, depending on your needs and preferences. However, it's important to note that machine learning models are not infallible and can be fooled by new or sophisticated deepfake methods.

4 Human verification

The final way to detect deepfakes is to rely on human verification, which involves contacting the original source, author, or subject of the image or video and asking them to confirm or deny its authenticity. To do this, you can reach out to the person or organization who published or shared the media file and ask for more information or evidence. You can also contact the person or entity that appears in the media and ask for their consent or opinion. Additionally, you can consult with experts, colleagues, or fact-checkers who have more knowledge or experience on the topic or technique. Tools like Whois, TinEye, and Google Reverse Image Search can help you find contact details, origin, or similar images of a media file.

5 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?