Deepfakes are AI-created or AI-edited videos, images, or audio clips that make people appear to say or do things they never actually did. As these fakes become more realistic on social media, it’s important to know how to recognize them.

Some common warning signs include unnatural facial expressions, poor lip-syncing, unusual blinking, or skin that looks overly smooth. You may also notice inconsistencies in lighting, shadows, or reflections. In videos, the edges around the face or hair can sometimes flicker or blur, especially during quick movements.

Audio can also reveal manipulation. AI-generated voices may sound robotic, emotionally flat, or have awkward pauses and inconsistent background noise.

Beyond visual clues, consider the context of the content. Deepfakes are often designed to provoke strong emotions such as shock, anger, or fear. Be cautious of sensational clips from unfamiliar accounts, especially if there is no clear source or full-length version available.

To verify suspicious content, you can reverse-search screenshots using tools like Google Images or TinEye. Fact-checking websites such as Snopes and Reuters Fact Check can also help confirm whether a video or image is authentic.

Although social media platforms try to limit harmful manipulated media, misleading deepfakes can still spread quickly. A good habit is to avoid trusting or sharing emotionally charged content until it has been verified by reliable sources.