What is a deepfake?
A deepfake is a video, image or audio recording altered by artificial intelligence to make it look like a person said or did something they never said or did. The technology uses deep learning to produce synthetic content that is extremely realistic and hard to tell from the real thing.
Threats to your privacy
Deepfakes can be used to create non-consensual pornography, impersonate you, put compromising words in your mouth, or damage your reputation. Anyone with enough photos or videos of you, easy to find on social media, could create a deepfake of you.
Social and political impact
Beyond individual privacy, deepfakes put society at risk: large-scale disinformation, election manipulation, believable fake news, and a loss of trust in visual media. It is getting harder to tell real from fake in digital content.
Detecting deepfakes
Several signs can give a deepfake away: unnatural blinking, lips out of sync with the audio, inconsistent lighting, and visual artifacts around the face. Automatic detection tools that use AI are in development, but it is a constant race between deepfake creators and detectors.
Protecting yourself from deepfakes
Limit the public photos and videos of you online. Use strict privacy settings on social networks. If you become a victim, document the deepfake, report it to the platform, and talk to a lawyer. New laws are being developed to criminalize the malicious creation of deepfakes.
In the age of deepfakes, seeing is no longer believing. Critical thinking matters more than ever.