What is a deepfake in AI, and what risks does it pose?

When splitting name 'deepfake' into 'deep'and 'fake' then 'deep' comes from 'deep learning' which is a advanced type of artificial intelligence (AI) that uses layered computer algorithms to process information, while 'fake' refers to content that is fabricated and untrue.

Deepfake AI is then synthetic media—such as video, audio, or images—generated by machine learning models like generative adversarial networks (GANs) or diffusion systems to impersonate a real person’s face, voice, or likeness without their consent.

Unlike traditional editing, these tools synthesize novel content that makes individuals appear to say or do things they never did, creating hyper-realistic fabrications that are increasingly difficult for humans to distinguish from reality.

Bad actors use voice cloning (requiring as little as three seconds of audio) and real-time video impersonation to deceive employees into authorizing fraudulent wire transfers or bypassing identity verification systems. Attackers leverage deepfakes to defeat biometric authentication controls, enabling them to steal identities, open accounts, or access sensitive data under the guise of a legitimate user.

Fabricated content is used to spread disinformation, damage reputations through defamation, or manipulate public opinion and financial markets by depicting public figures or executives making false statements. The technology is frequently exploited to create non-consensual explicit content or compromising material, which is then used to extort victims by threatening to release the fabricated media.

The pages below are an collection showing how bad actors are filling deepfake to "suppress" us all to think like them or to take revenge, defame and to threaten. 

• Pages may consist of a YouTube video.