Technology is showing us today what ancient texts and scientists proposed long ago. Vedic philosophy and even Lord Krishna in the Bhagavad Gita point to the illusory nature of the world and the limitations of human perception in discerning the true essence of the universe. Albert Einstein famously said, “Reality is merely an illusion, albeit a very persistent one.” Taking the form of a persistent, ever-growing illusion today are deepfakes.
Deepfakes are merely the combination of powerful deep learning algorithms and rapid advances in computing power leading to the emergence of a new generation of hyper-realistic synthetic media, including all forms of digital content — video, text, images, or audio — that have been either manipulated or fabricated using sophisticated algorithms to mislead, deceive, influence audiences or reinforce behavioural patterns and stereotypes.
Deepfakes are often used to replace a person’s likeness or voice. For instance, there have been cases in which eminent politicians, celebrities, and other leading personalities were impersonated to propagate some misleading views that they had never been involved with. Many have emerged from artificial intelligence (AI) platforms that have the potential to convert any text to a speech or a video in a few seconds, thereby creating a saga of deceit.
Also Read Bumps on the road: New BoT terms for highway construction may throw pvt investors into risk-aversion mode E-commerce needs a bulwark The elephant in RBI’s room The giant has arrived: India’s rise as a global power creates a tightrope of domestic and international aspects Impact on democracy
In November 2023, Prime Minister Narendra Modi said in a speech: “A new crisis is emerging due to deepfakes produced through artificial intelligence.” He also said that the misuse of AI to create deepfakes was problematic and that it would turn into a formidable problem.
The 2024 general elections in India, one among 40 polls globally, could see a massive explosion of deepfakes, posing a threat to the electoral integrity of the democratic process. Deepfakes featuring the voices and actions of politicians, along with fabricated speeches, will lead to misinformation, creating confusion among voters, and impacting their decision-making. A report by Al Jazeera said that nearly 40 campaign managers are looking for AI experts to create deepfakes and influence voters through WhatsApp groups. The growing phenomenon of deepfakes is a significant leap from earlier versions of media manipulation that were time-consuming and demanded expertise in creating fake content. Today, deepfakes are growing exponentially owing to four primary reasons:
1. The consistency and volume of media-generated deepfakes are unmatched because of sophisticated AI techniques that make it impossible to decipher synthetic and original content. For instance, fraudulent loan applications use AI tools to generate misleading celebrity endorsements/advertisements.
2. Deepfakes can fabricate digital content to create and circulate new material that did not exist before.
3. The ready availability and user-friendliness of deepfake tools is concerning. According to a news report, there has been a ten-fold increase in the number of complaints relating to morphed/doctored images and videos online.
4. The anonymity of users coupled with the easy availability of distribution channels, particularly on social media, enables faster propagation of deepfake content.
In India, any content can go viral due to low latency, and ease of access to millions of smartphones and social media accounts can potentially cause irreparable damage to the dignity or image of any personality. Regulators are hugely concerned and experts are increasingly advising the deployment of a ‘zero trust architecture’.
Way forward: A multi-stakeholder approach
Given the magnitude of the problem, combating deepfakes would require a multi-stakeholder approach — a mix of regulations, innovative solutions by companies, and user-awareness/fact-checking initiatives by civil society organisations. While India does not have a specific law on deepfakes, the government has indicated that existing regulations, including provisions of the Information Technology (IT) Act, 2000, and IT Rules, 2021, deal with this issue with adequate relief provided through punitive actions, penalties for offences, content moderation, strict content takedown timelines, etc. A more principle-based approach to combat the harms of AI is expected in the upcoming Digital India Act. Moreover, 20 companies (including Meta, Google, OpenAI, and Microsoft) are finalising a pledge, also known as the Munich accord, to fight back against the deceptive use of AI in the 2024 elections. Simultaneously, several bigtech companies are developing software solutions that help in the watermarking and labelling of deepfake content.
Experts feel that competing to combat deepfakes is a race to the bottom and that a single solution could be inefficient to deal with the issue. With more collaboration among players, regulatory guidance, detection tools, and user awareness, we can attempt to ensure that reality remains grounded in truth and well distinguished from illusion.
TV Ramachandran, Honorary fellow, IET (London) and president, Broadband India Forum with research inputs by Mira Swaminathan, BIF; Views are personal
Technology is showing us today what ancient texts and scientists proposed long ago. Vedic philosophy and even Lord Krishna in the Bhagavad Gita point to the illusory nature of the world and the limitations of human perception in discerning the true essence of the universe. Albert Einstein famously said, “Reality is merely an illusion, albeit a very persistent one.” Taking the form of a persistent, ever-growing illusion today are deepfakes.
Deepfakes are merely the combination of powerful deep learning algorithms and rapid advances in computing power leading to the emergence of a new generation of hyper-realistic synthetic media, including all forms of digital content — video, text, images, or audio — that have been either manipulated or fabricated using sophisticated algorithms to mislead, deceive, influence audiences or reinforce behavioural patterns and stereotypes.
Deepfakes are often used to replace a person’s likeness or voice. For instance, there have been cases in which eminent politicians, celebrities, and other leading personalities were impersonated to propagate some misleading views that they had never been involved with. Many have emerged from artificial intelligence (AI) platforms that have the potential to convert any text to a speech or a video in a few seconds, thereby creating a saga of deceit.
In November 2023, Prime Minister Narendra Modi said in a speech: “A new crisis is emerging due to deepfakes produced through artificial intelligence.” He also said that the misuse of AI to create deepfakes was problematic and that it would turn into a formidable problem.
The 2024 general elections in India, one among 40 polls globally, could see a massive explosion of deepfakes, posing a threat to the electoral integrity of the democratic process. Deepfakes featuring the voices and actions of politicians, along with fabricated speeches, will lead to misinformation, creating confusion among voters, and impacting their decision-making. A report by Al Jazeera said that nearly 40 campaign managers are looking for AI experts to create deepfakes and influence voters through WhatsApp groups. The growing phenomenon of deepfakes is a significant leap from earlier versions of media manipulation that were time-consuming and demanded expertise in creating fake content. Today, deepfakes are growing exponentially owing to four primary reasons:
1. The consistency and volume of media-generated deepfakes are unmatched because of sophisticated AI techniques that make it impossible to decipher synthetic and original content. For instance, fraudulent loan applications use AI tools to generate misleading celebrity endorsements/advertisements.
2. Deepfakes can fabricate digital content to create and circulate new material that did not exist before.
3. The ready availability and user-friendliness of deepfake tools is concerning. According to a news report, there has been a ten-fold increase in the number of complaints relating to morphed/doctored images and videos online.
4. The anonymity of users coupled with the easy availability of distribution channels, particularly on social media, enables faster propagation of deepfake content.
In India, any content can go viral due to low latency, and ease of access to millions of smartphones and social media accounts can potentially cause irreparable damage to the dignity or image of any personality. Regulators are hugely concerned and experts are increasingly advising the deployment of a ‘zero trust architecture’.
Given the magnitude of the problem, combating deepfakes would require a multi-stakeholder approach — a mix of regulations, innovative solutions by companies, and user-awareness/fact-checking initiatives by civil society organisations. While India does not have a specific law on deepfakes, the government has indicated that existing regulations, including provisions of the Information Technology (IT) Act, 2000, and IT Rules, 2021, deal with this issue with adequate relief provided through punitive actions, penalties for offences, content moderation, strict content takedown timelines, etc. A more principle-based approach to combat the harms of AI is expected in the upcoming Digital India Act. Moreover, 20 companies (including Meta, Google, OpenAI, and Microsoft) are finalising a pledge, also known as the Munich accord, to fight back against the deceptive use of AI in the 2024 elections. Simultaneously, several bigtech companies are developing software solutions that help in the watermarking and labelling of deepfake content.
Experts feel that competing to combat deepfakes is a race to the bottom and that a single solution could be inefficient to deal with the issue. With more collaboration among players, regulatory guidance, detection tools, and user awareness, we can attempt to ensure that reality remains grounded in truth and well distinguished from illusion.
TV Ramachandran, Honorary fellow, IET (London) and president, Broadband India Forum with research inputs by Mira Swaminathan, BIF; Views are personal
Get live Share Market updates, Stock Market Quotes, and the latest India News and business news on Financial Express. Download the Financial Express App for the latest finance news.
login
who are we?
contact us
qosheapp
Toi Staff
Gideon Levy
Belen Fernandez
Mort Laitner
Rami G Khouri
Ali Fathollah-Nejad
Nikkei Editorial
visit website
