An AI-generated video in which Kerala Chief Minister V D Satheesan is seen dancing along with Kerala State Electricity Board employees has created a brouhaha with implications far beyond the state. By the time Instagram owner Meta pledged on Monday to restrict access to the video, it had been seen by crores of users; copies of it are still attracting views online. It has been shared at a time the state is reeling under a severe power shortage.
Satheesan declared his surprise and shock at the near-perfect rendition in the artificially-created video. He even made light of it, saying his AI avatar danced better than actor Mohanlal and Tamil Nadu Chief Minister C Joseph Vijay. Jokes apart, the incident refocuses attention on an increasingly pertinent question: are Indian laws keeping up with ever-evolving AI usage?
The Kerala CM is the latest in a long line of Indian public figures whose deepfakes have confused, entertained and enraged viewers at the same time. It did not help when some official X handles of the BJP and the Congress recently shared objectionable morphed images of each other’s leaders. Such actions, much like Donald Trump sharing his own deepfakes, validate felonious behaviour to credulous netizens.
Though the legal framework addressing deepfakes has evolved, including rules promulgated earlier this year mandating their clear labelling and faster takedown, implementation remains constrained. While Section 66C of the Information Technology Act, 2000 punishes identity theft, Section 66D punishes cheating by digital personation. To be considered a crime, both require intent to defraud or an act of cheating. Thus, the law mainly focuses on punishing the consequences of deepfakes without directly criminalising the initial act of using someone’s face, voice or identity without consent. Mischief and misinformation can flourish in the gap.
An opportunity to change course is coming up at the Supreme Court, which has recently issued notice to transfer all deepfake regulation cases across the country to itself. Non-consensual or unauthorised creation of replicas of another person’s image, voice, or any biometric identity through AI must be criminalised irrespective of what’s being done with it. The legal approach must be more preventive in nature than reactive. It must be recognised that the theft of identity is wrongful in itself. As AI evolves at warp speed, it is imperative for India’s legal system to keep apace.