ChatGPT and Amazon are yielding an entire new breed to writers. Literally. No longer does one have to be prolific with ideas and thrifty with words. As long as you have a laptop and an AI-empowered browser, becoming a writer is not that difficult anymore.
Now, taking the AI altered norms a bit further, comes the emerging threat of AI-powered deepfake videos. This is particularly true in a country like India, which is known for its volatile turn of events against any political backdrop.
It may be recalled that Chief Election Commissioner Rajiv Kumar, on January 23, flagged the disturbing trend of “deep fake narratives” that had become a common feature in elections worldwide. Kumar was speaking at a three-day international conference on the “Use of Technology & Elections Integrity” organized by the Election Commission of India in Delhi.
He advocated pro-active moves by social media intermediaries to detect deep fakes proactively. “This would serve well especially in the Indian context where electoral cycles are definite and scheduled well in advance,” he stated, adding that the Indian police does not have competitive tools to detect deepfake videos.
Law enforcement agencies in Gujarat have sounded an alert about the threat of AI-simulated images and clips. The police have already encountered two instances where manipulated videos of politicians were released in the state.
While one was reported from Surat in April 2021, the other came to light in November 2022 during the peak of political campaigning for the polls held in December.
According to records, the cybercrime police station in Shahibaug was tipped off by Pravin Chandrana, the president of the legal cell of AAP Gujarat, on November 11, 2022. The complaint noted that some local news channel circulated an obscene video of their party leader and Delhi MLA, Gulab Singh Yadav, ahead of the 2022 Assembly polls.
“It is clear that Yadav’s face has been superimposed on someone else’s body. Someone used some editing app to create the obscene clip. The person in the video is not Yadav,” the complaint had stated.
The second case involved former chief minister Vijay Rupani. The Surat crime branch arrested embroidery businessman Kishan Arvind Rupani from Dhandhuka allegedly for circulating a deepfake video of Rupani on social media. In the video, Rupani was seen singing the Taylor Swift number, “I knew you were trouble.”
The 28-year-old businessman Kishan Arvind Rupani from Dhandhuka told the police he did it to gain more followers and “likes” on social media. Kishan was charged and is awaiting trial, said a Surat crime branch official.
Deepfakes first gained attention in 2017 when the technology was used to create fake porn videos using celebrity footage and AI algorithms.
On the political front, the technology was first used by a comedian impersonating Barack Obama to deliver a public service announcement (PSA) video on how deepfakes can be deceptive. Such videos can be created for as little as (Rs 16,500 to Rs 22,000 approx).
“There is no way to prevent this. With AI bots now openly accessible and legal, the menace will assume manifold proportions. Mischief makers just need a sample of your voice and a clip of your video to replicate facial features and characteristics. Making matters worse is the fact that it is difficult to trace the digital footprint of such videos,” shared an official from the Gandhinagar FSL.
The problem is that the videos seem very real. “By the time a person registers a complaint and the police begin an investigation, the victim’s reputation has been sullied. In the context of election campaigns, it is not tough for someone to fake a video to spread lies or issue threats to foment trouble. We worry we will have to deal with deepfake videos as 2024 election campaigning heats up.”
Also Read: Ecological Wellness Behind Crocs On The Rise In Charotar Wetlands
