Deepfake technology, which involves creating synthetic media that portrays events or images that never actually occurred, has a relatively recent history. The term "deepfake" is a combination of "deep learning" and "fake," and it refers to the use of artificial intelligence (AI) to generate convincing fake content.
The concept of deepfake became widely known in 2017 when a Reddit user created a subreddit dedicated to sharing videos that used face-swapping technology to insert celebrities' likenesses into existing videos, often for pornographic purposes. This use of AI for creating realistic-looking media quickly raised concerns about its potential for misuse, particularly in the creation of fake news, hoaxes, and other forms of disinformation.
Deepfakes are produced using two main AI algorithms: one that creates a synthetic image or video, and another that detects whether the replica is fake. The creation algorithm adjusts the synthetic media based on feedback from the detection algorithm until it becomes indistinguishable from real media.
The technology behind deepfakes has evolved from earlier forms of media manipulation, with photo manipulation dating back to the 19th century and applied to motion pictures as technology improved. However, the rapid advancement of AI in the late 20th and early 21st centuries has made deepfakes much more accessible and difficult to detect.
The history of deepfake technology is quite fascinating and involves a mix of academic research and community-driven development. Here's a brief overview:
Early Development: The foundations of deepfake technology can be traced back to the 1990s, with researchers at academic institutions exploring the potential of AI in media manipulation.
Generative Adversarial Networks (GANs):
A significant leap in the technology came with the invention of GANs in 2014 by computer scientist Ian Goodfellow. GANs are a class of AI algorithms used in unsupervised machine learning, implemented by a system of two neural networks contesting with each other in a zero-sum game framework.
Public Emergence: The term "deepfake" emerged in 2017 when a Reddit user, known by the pseudonym 'deepfakes', began sharing videos on a subreddit that used machine learning to swap celebrities' faces onto existing videos, often for pornographic content.
Widespread Attention: This use of AI caught public attention and raised concerns about its potential for creating convincing fake content that could be used for disinformation or other malicious purposes.
Creators involved
The inventors/creators involved in the development of deepfake technology have ranged from academic researchers to anonymous online community members. The technology has since evolved, becoming more accessible and sophisticated, leading to a wide range of applications beyond the initial controversial uses. As deepfake technology continues to develop, there is an ongoing discussion about its ethical implications and the need for regulations to prevent misuse.
Deepfake technology has led to the creation of various notable projects and the emergence of skilled creators. Here are some of the most prominent examples:
EZ RyderX47: This creator gained fame for a deepfake video where Marty McFly and Emmett "Doc" Brown from "Back to the Future" are replaced by Tom Holland and Robert Downey Jr., respectively. The video showcases the creative possibilities of deepfake technology.
McAfee's Project Mockingbird: Announced at CES 2024, this project aims to empower users to identify deepfakes. It gained attention when it was used to debunk a deepfake scam involving a fake Taylor Swift promoting cookware.
McAfee Advisory! No, That’s Not Taylor Swift Promoting Le Creuset Cookware.
— McAfee (@McAfee) January 10, 2024
If you see this video in your social media feed, we can confirm that it is a #deepfake scam generated through #AI.
McAfee's Project Mockingbird technology announced at #CES2024, is designed to empower… pic.twitter.com/aVpcjI0Pgk
In Event of Moon Disaster: This short film features an incredibly convincing deepfake of Richard Nixon delivering a speech that was prepared in case the Apollo 11 mission had failed. The film explores the implications of deepfakes and their historical context.
These projects and creators have contributed to both the advancement of deepfake technology and the ongoing conversation about its ethical use and societal impact.
Losses incurred due to deepfake tech
Deepfake technology has led to significant financial and social losses over the years. Some of the notable impacts are as :Financial Losses: Deepfake scams have resulted in losses ranging from $243,000 to $35 million in individual cases. For instance, a bank manager was tricked into transferring $35 million to a fraudulent account due to a deepfake audio message.
Business Impact: A report from 2020 projected that deepfakes could cost businesses globally $250 billion by 2025. Financial institutions might face annual losses of up to $30 billion due to deepfake fraud by 2027.
Cybersecurity Threats: In 2022, 66% of cybersecurity professionals experienced deepfake attacks within their organizations. The banking sector is particularly concerned, with 92% of cyber practitioners worried about its fraudulent misuse.
Social Engineering Attacks: Deepfakes have been used to create fake videos or audio messages, often impersonating CEOs or other high-ranking executives to deceive individuals into sending money or disclosing sensitive information.
Misinformation and Public Trust: Deepfakes have the potential to undermine election outcomes, social stability, and even national security, especially in the context of disinformation campaigns.
Donald Trump Case: In 2018, a deepfake video of Donald Trump was released by a Belgian political party, urging Belgium to withdraw from the Paris climate agreement. Although intended as satire, it highlighted the ease of manipulating a world leader's image.
Trump heeft een boodschap voor alle Belgen... #Klimaatpetitie pic.twitter.com/Kf7nIaDOKj
— Vooruit (@vooruit_nu) May 20, 2018
Deepfake Voice Scam: In 2019, criminals used deepfake technology to mimic a CEO's voice in a fraudulent attempt to transfer funds, showcasing the potential for financial scams.
The rise of deepfake content and its misuse has prompted discussions on the need for better regulation and detection technologies to combat this issue and mitigate its harmful effects.
Criminalization of deepfake technology misuse
The issue of deepfake technology and its criminalization is a complex and evolving area of law globally, including in India. While deepfakes have potential benefits in various fields, they also pose significant risks such as privacy violations, defamation, and the spread of misinformation.Globally, there is a growing concern about the malicious use of deepfakes, and countries are exploring ways to regulate this technology. The legal status of tackling crimes related to deepfakes varies from country to country, with some having specific regulations while others rely on existing laws to address the issue.
In India, as of the information available up to 2021, there was no specific statute that directly addressed deepfake cybercrime. However, various other laws could be applied to combat crimes involving deepfakes. For instance, Section 66E of the Information Technology (IT) Act of 2000 could be invoked in cases of deepfake offenses that infringe on a person's privacy by capturing, publishing, or transmitting their image without consent.
Experts have pointed out that while India and other countries face challenges due to deepfakes, practical solutions are available, and provisions under several pieces of legislation could offer both civil and criminal relief. It's important to note that the development and use of deepfakes is a global issue, likely requiring international cooperation to effectively regulate their use and prevent associated crimes.
Advertisements