What are deepfakes and how they are dangerous

Deepfakes refer to manipulated videos or other digital depictions created by sophisticated artificial intelligence that produce real-life images and sounds.
A system of deep learning can produce a convincing fake by studying a target person's photographs and video from multiple angles, and can then imitate their behavior and speech patterns.
If someone wishes to act maliciously towards someone who has authority, the results could be as simple as creating a fake video that damages their reputation. Videos can reach a global audience in a matter of minutes and spread over the Internet. In such situations it is extremely difficult for a company to monitor damage by this level of reach and speed.
The examples here show why people can't believe what they see or hear anymore. Although this coverage focuses on fake videos, similarly credible examples of pictures exist. Media can improve an understanding of a person's problem and clarify points which might otherwise mislead. But everyday people can easily access sophisticated software that enables them to communicate the things they want to doctor media.
This shift in the media landscape means that people have to check the facts and verify the sources instead of automatically believing that what is in front of them is true. This change will take time, but the risk of deep-fakes and their part in stopping the spread of them is crucial for society members to understand.
Large corporations such as Facebook and Microsoft have taken initiatives to detect and remove deep-fake videos. According to Reuters, both companies have previously announced that they will cooperate with top universities throughout the US to create a large database of fake videos for research.