Why is it in the News?

The government has warned top social media and internet companies that their platforms may be temporarily suspended and even be ordered blocked in case they are unable to tackle the menace of deepfakes.

What are Deepfakes?

• The term "deepfake" combines the concepts of deep learning with the fabrication of content.
• Deepfakes involve the creation of synthetic images and audio using machine-learning algorithms, intending to disseminate misleading content by replacing a real person's appearance, voice, or both with artificially generated likenesses or voices.
• These manipulated creations can either depict nonexistent individuals or simulate real people engaging in actions or utterances they never did.
• Originating in 2017, the word "deepfake" emerged when a Reddit user named "deepfakes" shared explicit videos featuring celebrities.
• The process of crafting deepfakes utilizes machine learning models employing neural networks to manipulate visual and auditory elements.
• To generate a convincing deepfake video, creators train the neural network on extensive real footage of the targeted person, facilitating a realistic understanding of their appearance from various angles and lighting conditions.
• This trained network is then combined with computer graphics to overlay the person onto a different actor.
• Regrettably, this technology is increasingly exploited for malicious purposes, including scams, celebrity impersonation, election interference, social engineering, disinformation attacks, identity theft, and financial fraud.
• The distinguishing factor of deepfakes lies in their challenging detection due to their sophisticated nature.