As political parties are all out to woo voters ahead of the Lok Sabha elections 2024, cybersecurity experts have raised concerns over the possible use of artificial intelligence and deepfake technology to influence the electorate and mislead the public.
Prime Minister Narendra Modi, in a recent conversation with Microsoft cofounder Bill Gates, highlighted the dangers of deepfake videos, saying important measures should be taken while using AI in the tech era.
“They can use my voice and it can deceive people initially,” he stressed on the need for global do’s and don’ts to use this technology.
“During election season, the development of AI technologies such as deepfake videos, and voice cloning pose complex challenges to the integrity of the electoral process,” Mohan Subrahmanya, Country Leader, Insight – India told LiveMint.
Hence, the government along with cybersecurity experts and technology providers should work together to protect democratic institutions from such potential AI threats, he said.
Pawan Prabhat, cofounder of ShortHills AI believes that the targeted deep fakes of political leaders are more difficult to identify as true or false and can significantly influence public perception, potentially impacting the outcome of upcoming Lok Sabha elections.
Subrahmanya said that regular security testing, sophisticated detection technology, and robust public education initiatives with transparency guidelines and international cooperation will help distinguish authentic content from deepfakes and prevent political leaders from becoming victims of digital fraud.
Citing examples of OpenAI and Meta, he said companies in the United States have included watermarks for AI-generated photos to facilitate identification while the University of Maryland use QR codes to verify legitimate live recordings.
Prabhat also suggested several standard ways to determine whether a video is original or AI-generated. “Websites such as Detect Deepfakes by the MIT Media Lab and other academic institutions offer tools and knowledge to help identify common visual cues, like the appearance of the eyes, brows, shadows, and reflections.”
In India, the Indian Cyber Crime Coordination Center (I4C) has been designated as the nodal agency to deal with objectionable online content, while YouTube‘s new policy allow for the removal of misleading AI-simulated content.
However, the ShortHills AI asserted that there is no simple solution to the AI threat, even in other countries, the response has been to call for tighter regulation and the development of concrete policies, along with severe punishments for the creation and distribution of deepfake content.
“The US government has already enacted some relevant laws, but the Indian government has yet to pass comprehensive legislation on this issue. Given the proximity of the upcoming elections, a more immediate solution could be to implement a strict, blanket ban on deepfake content until the elections are over,” Prabhat said.
It is important to note that the greater concern is not about deepfake videos targeting famous political personalities, but the highly personalized deepfake videos or audios aimed at specific groups of people can have a significant impact on election results, he added.
All 543 constituencies in India will go to polls in seven phases beginning from April 19 and concluding on June 1. The counting of votes will take place on June 4.
Unlock a world of Benefits! From insightful newsletters to real-time stock tracking, breaking news and a personalized newsfeed – it’s all here, just a click away! Login Now!
The Mint News App to get Daily Market Updates & Live Business News.
Published: 08 Apr 2024, 02:01 PM IST
