Concept of deepfakes with real-life examples
Deepfakes represent the manipulation of images or videos through the usage of artificial intelligence (AI) in which the face of an individual in an existing image or video is replaced with the face of someone else. This becomes possible because of the emergence of generative AI, which supports the generation of new content on the basis of available data through the usage of algorithms.
In recent years, major attention has been gained by this technology because of its effectiveness in manipulating and deceiving viewers. Around 500,000 of this type of voice and videos were circulated across the social media platform worldwide. Considering the current trends, there is the potential for such types of online content to reach 8 million in 2025.
This term came in the year of 2017 when a Reddit user in the name of deepfakes was found to be popularizing the technique through the formation and posting of explicit content by swapping the faces of celebrities with existing pornographic content. After that, the accessibility of this technology has considerably increased, leading to the enhancement of concerns regarding their multifaceted impacts on society.
Some of the prominent examples of the fake content that has been spread extensively include an image of the former president of the US, Donald Trump, fighting with police, an image of Pope Francis wearing a puffer jacket, a dancing video of Queen Elizabeth describing regarding the potential of technology as well as a video in which Mark Zuckerberg, the Facebook CEO said about the offensive power of his company. The occurrence of these events has not been seen in real life.
The potential influence of deepfakes on society
There is a significant power of fake content to pose an adverse impact on individuals as well as society. The sharing of fake news and misinformation is one of the major concerns related to the increasing use of this technology. This has the capability the manipulation public opinion as well as harm the reputation of institutions or public figures through the formation of realistic videos or images involving such activities they had never done.
There can be the potential usage of this technology to defame persons by appearing them in offensive situations or doing wrong activities. Many individuals and female celebrities are found to be the victims of these misleading activities where their faces are swapped onto explicit content and circulated across digital media.
In this regard, the case of Taylor Swift can be highlighted, whose digital photograph has been used for the creation and sharing of pornographic content on the social media handle X (Previously referred to as Twitter). As a result of this case, Congress and the fans of Taylor Swift have expressed their indignation as well, and the conversation regarding digital regulations and rights has been increased.
Towards addressing the fake images of Taylor Swift, X was found to be removing all these images and undertaking actions for the accounts accountable for sharing them. Moreover, the users were asked to report the accounts. However, it took a lot of time, which led millions of users to see the images. This proactive approach to social media is not sufficient. Rather, strict initiatives should be implemented as this abuse does not only breach privacy but also harm psychological and emotional well-being.
Posting Non-Consensual Nudity (NCN) images is strictly prohibited on X and we have a zero-tolerance policy towards such content. Our teams are actively removing all identified images and taking appropriate actions against the accounts responsible for posting them. We're closely…
— Safety (@Safety) January 26, 2024Usage of deepfakes in business
This technology has been seen to be utilised in the business for both legal and illegal purposes. Concerning the illegal aspects, it has been used for the purpose of making frauds. For instance, recently, a finance worker at a multinational company was found to be paying 25 million US dollars to deceivers who were engaged in the usage of deepfake technology to look like the chief financial officer and other staff of the firm in a conference video call. While at first, it seemed to be suspicious, the finance officer became the victim of this fraud by proceeding towards the money transfer as the individuals involved in the video call seemed like known employees of his organisation.
On the contrary, pertaining to legitimate business purposes, its usage has been observed in the extension of marketing reach and influence of a seller. For example, deepfakes of Chinese influencers are found to be used for live streaming for the promotion of products on the popular e-commerce platform in China like Taobao.
In recent times, live streaming has become one of the prominent ways of promoting the products for digital and traditional brands in China. In Taobao, Kuaishou, Douyin and other platforms, influencers are seen accountable for selling goods worth more than one billion dollars in one night.
However, there is a significant cost associated with the retention of livestream hosts, offering training to them as well as identifying technical insights of broadcasting, which are a challenging task for smaller brands. On the other hand, it is quite inexpensive to automate these tasks with the formation of deepfake videos, which can be used for live streaming 24/7. The operating costs of a company can be minimized through the usage of the likeliness of online streamers 24/7.
Ethical consideration for the formation and utilisation of deepfakes
The formation and sharing of an individual’s likeness without consent leads to the violation of the ethical rights to privacy and autonomy of the individual, as well as promotes misrepresentations. Without the consideration of the ethical guidelines stated below, the effect of fake content can be wider and can mislead the understanding of reality and truth.
Purpose and intention
There should be a legitimate intention for designing and employing deepfakes for lawful research or application in education. On the other hand, the intention of deceiving or privacy breaches, such as sharing misinformation, developing offensive content, or imitating individuals for the purpose of committing fraud, should be prevented.
Consent
Undertaking of consent must be required from the persons before creating and using the deepfakes of them. The individuals should be informed clearly how their generated images, videos and voices will be utilised along with the purpose of the usage. Additionally, an ongoing consent procedure should be involved rather than a one-time deal. Individuals should be empowered the withdrawal their consent in case of the occurrence of such a situation when they perceive that their likeness is being utilised in the wrong way.
Transparency
It is quite difficult for the general public to differentiate fake content from genuine content. Therefore, there is the requirement of the maintenance of sufficient transparency in the form of adopting disclosure in the case of swapping the face, especially when public opinion can be considerably influenced by the authenticity of the contents. Furthermore, for the implementation of transparency on a wider scale, increasing verification of the authenticity of the sources of online content should be required.
Mitigation ways of the rising deepfakes attacks
Investment in technology
The advanced technologies have the potential for the identification and detection of fake content as soon as possible when they occur. Therefore, businesses should leverage efficient detection tools to prevent malicious content irrespective of any format, including images, videos or audio.
Arrangement of training programs
The companies are required to offer comprehensive training programs to their workforce not only to raise awareness regarding the societal, legal and ethical issues of deep fakes but also to develop the essential skills and competencies among them for the recognition and prevention of those content immediately. Not only the employees but it is also crucial to raise awareness among the public with the essential insights and tools for identifying and responding to this issue. Therefore, there is a requirement for the inclusion of media literacy programs in the school curriculum along with the arrangement of public awareness campaigns.
Regular audits
Considering its nature of posing a threat to data privacy, the conduct of regular security audits is vital for finding and mitigating the risks across communication channels. Implementation of robust security protocol along with regular upgradation is one of the major preventive measures to the cybersecurity and data privacy threat.
International regulations and standards
In order to address the issues posed by unwanted and offensive content, governments, international regulatory bodies, and technology organisations should work collaboratively for the establishment of efficient regulations and guidelines for the ethical usage of AI technologies.
Initiatives of the government of India related to deepfakes
The government of India has issued a set of guidelines with the introduction of IT Rules 2021 under the IT Act, 2000. According to these rules, the intermediaries in social media are accountable for the withdrawal of deepfake photos and videos actively, and failure to perform this activity can lead to a penalty of Rs 1 lakh and even imprisonment of up to three years. This rule is enforced from May 2026 to 2021. Moreover, according to Rule 3(2)(b), such contents on which complaints have been registered should be erased within 24 hours of the complaint.
FAQ
What are deepfakes?
Deepfakes represent the manipulation of images or videos through the usage of artificial intelligence (AI) in which the face of an individual in an existing image or video is replaced with the face of someone else.
When did the term deepfakes first appear?
The term deepfakes came in the year of 2017 when a Reddit user by the name of deepfakes was found to be popularising the technique through the formation and posting of explicit content through swapping of the faces of celebrities with the existing pornographic content.
How many deepfakes were shared on social media in 2023?
Around 500,000 deepfake voices and videos were circulated across social media platforms worldwide.
What is the potential influence of deepfakes on society?
Deepfakes have the capability the manipulation public opinion as well as harm the reputation of institutions or public figures through the formation of their realistic videos or images involving such activities they had never done.
What are some examples of deepfakes?
Some of the prominent examples of the deepfakes that have been spread extensively include an image of the former president of the US, Donald Trump, fighting with police and an image of Pope Francis wearing a puffer jacket.
