On October 1st, actor Tom Hanks took to Instagram to alert his followers about a concerning case of deepfake usage. He discovered that a dental plans advertisement was using a fake version of himself. He posted a message to clarify, saying, "BEWARE!! There's a video out there promoting some dental plan with an AI version of me. I have nothing to do with it."
On November 6, Indian actress Rashmika Mandanna took to Instagram to call out a viral deepfake video of herself. Her face had been manipulated onto an Instagram video posted by a British-Indian influencer named Zara Patel.
Soon after that, a manipulated image of Katrina Kaif from her upcoming film 'Tiger 3' also began circulating. Hollywood actress Scarlett Johansson also faced deepfake challenges when her face was morphed into adult films. Actor Robert Pattinson became a target of deepfake videos posted on a parody TikTok account.
Unfortunately, the list does not end here. Celebrities have been targets of such manipulation for months, raising concerns about the extent of damage these deepfake videos can cause. These crimes not only target celebrities but can impact anyone out there. With easy access to deepfake AI tools, anyone could become a victim.
What exactly is a deepfake?
Deepfake technology employs a form of artificial intelligence known as deep learning to generate counterfeit images of events that never transpired. It is often used to digitally alter a person's appearance and create deceptive video content.
Is it limited to just videos?
No, deepfake technology extends beyond videos. It can generate entirely fake but highly convincing photos from scratch. Furthermore, deepfake audio is increasingly prevalent, producing very convincing audio fakes.
What does the law say about deepfake crimes?
Currently, there aren't specific laws for deepfake crimes, but existing laws like the Information Technology Act offer some protection. If a deepfake video invades someone's privacy or impersonates them, they can file a complaint. Section 66D of the IT Act deals with the punishment for cheating by personation using a computer resource. Offenders can face up to three years in prison and a fine of up to Rs 1 lakh if found guilty. Under Section 66E of the IT Act, if someone intentionally captures, publishes, or shares an image of a person without their consent, and this violates their privacy, it can lead to imprisonment of up to three years or a fine of up to 2 lakh rupees or both. If a deepfake video aims to harm someone's reputation with false information, the affected person can file a defamation lawsuit against the creator.
How do you spot a deepfake?
Actually, it becomes more challenging as the technology improves. Poor-quality deepfakes are easier to identify, but highly professional and advanced ones are extremely difficult to distinguish from reality. Nevertheless, certain indicators may still be helpful. These indicators include:
• Inconsistent facial expressions and movements
• Unusual eye contact or gaze
• Audio quality and lip-syncing issues
• Inconsistent lighting and shadows
• Background artifacts or anomalies
• Abnormal blinking and breathing patterns
• Variations in video resolution and quality
• Cross-referencing with multiple sources
• Utilizing reverse image and video search tools
Are deepfakes inherently malicious?
The answer is no. While deepfake technology has been misused for fraudulent, deceptive, and harmful purposes, it also holds significant potential for constructive applications. Industries such as film and entertainment utilize deepfakes for special effects and historical figure re-creation. Education benefits from deepfakes for language-learning materials, reenactments, and virtual lectures. Deepfake technology is integral to various research areas, including computer vision, artificial intelligence, and human-computer interaction. Additionally, artists use deepfake tools for creative expression, generating artwork and music.
So, while deepfake technology itself is a tool with a wide range of potential applications, it's the intentions and actions of those who create and use deepfakes that determine whether they are malicious or not. Efforts are being made to develop tools and regulations to detect and mitigate malicious deepfake use while preserving legitimate and ethical uses.
