Urgency of Ai Verification Rules to Enforce the Validity of Information on the Deepfake Phenomenon: A Comparative Study of Indonesian Positive Law and the Blueprints for an Ai Bill of Rights of the United States

Electronic Journal of Education, Social Economics and Technology Vol. 6, No. 2, (2025), pp. 1~9, Article ID: 282 ISSN 2723-6250 (online) DOI: https://doi.org/10.33122/ejeset.v6i2.282 Research Article Urgency of Ai Verification Rules to Enforce the Validity of Information on the Deepfake Phenomenon: A Comparative Study of Indonesian Positive Law and the Blueprints for an Ai Bill of Rights of the United States Hanuun Zainum A*, Laina Rafianti Faculty of Law, Universitas Padjadjaran, Kampus Unpad Jatinangor, Jl. Ir. Soekarno KM. 21, Jatinangor, Sumedang, Jawa Barat, Indonesia *Corresponding Author: ABSTRACT Artificial intelligence technology, also known as human AI, is becoming more common and developing as human civilization becomes more modern. AI is not only known as humanoid robots; AI has also developed into software, computer systems, and programs. One of the AI innovations that has attracted attention is deepfake. The development of image and video manipulation technology has had serious consequences for the integrity of information and the reputation of agencies and individuals. This AI technology can easily cause disinformation. It is very easy to use this technology to spread misleading information, influence public opinion, and even trigger civil unrest. Furthermore, the author will compare Indonesian positive law and the Blueprints for an AI Bill of Rights as a guideline for United States AI to be able to minimize deepfake against disinformation. Legislation and conceptual theory are the main approaches used in the juridical-normative legal research method. Keywords: Artificial Intelligence; Deepfake; AI Guidelines 1. INTRODUCTION In line with the advancement of science, technological development is necessary as part of a country's development process. In the digital era, humans must utilize easy access to information and communication technology that supports their daily lives to continue creating a more advanced society. The overall lifestyle of modern humans is inseparable from electronic gadgets. In practice, theories and principles are often not in line with the reality on the ground and will always go hand in hand with unexpected possibilities. In the digital era, this poses new challenges for technology in human life. The development of artificial intelligence (AI) technology innovation is synonymous with Industrial Revolution 4.0 and Industrial Revolution 5.0. There is no denying that AI significantly impacts people's daily lives. The concept of the 5.0 era was first adopted by Japan in the Basic Policy on Economic and Fiscal Management and Reform 2016 through the 5th Science and Technology Basic Plan, aspiring to create a society where everyone can live a better life. Therefore, integrating the virtual and real worlds to develop data, information, and a higher quality of life is key to achieving this goal. In fact, according to the International Telecommunication Union (ITU), the UN body involved in telecommunications, artificial intelligence will impact most of our lives in the future (Ramli, 2024c). Artificial intelligence is defined by the World Intellectual Property Organisation (WIPO, 2016) as the creation of computer-based technologies and systems capable of performing tasks that require human intelligence through machine learning and deep learning. Thus, technologies that enable machines to act intelligently like humans are what we call artificial intelligence. The ability of robotic machines to learn from experience and perform activities associated with human intellectual abilities, including language comprehension, logic, and problem-solving, is known as artificial intelligence. This is inseparable from the impact of digital transformation as a consequence of technology that causes artificial intelligence to continue to develop and become more sophisticated (Kristophorus Hadiono, 2020). At the same time, the results of artificial intelligence cause the output of obscurity and are difficult to accept by Indonesian law because its existence is not supported by apparent legal certainty (Rizki Fauzi, Tasya Safiranita Ramli, 2022). Artificial Intelligence-Crime (hereinafter abbreviated as AIC) is a term for artificial intelligence misused by cybercriminals. The deepfake phenomenon or audio/video impersonation is one example of AIC in today's technological Page 1 of 9 Zainum A & Rafianti Electronic Journal of Education, Social Economic and Technology, Vol. 6, No. 2, (2025), pp.1~9, Article ID: 282 advancement. Deepfake technology has caused potential chaos in cyberspace and the real world. The term "deepfake" refers to an algorithm that allows users to mimic real-world visual objects by transforming one person's face into another. This technique can be applied to both photos and videos. Deepfake is an artificial intelligence-based method or approach to creating synthetic human images (Ellen Kusuma and Nenden Sekar Arum, 2022). Furthermore, the process behind deepfake involves two main methods, namely: a. Deep Neural Networks (DNN): With DNN, deepfake production using machine learning technology can blend a person's face into another video. Collecting the source video and target face data is the first step in this procedure. Deep learning methods train a network model to perfectly mimic eye, lip, and facial movements through extensive literacy and training. After training, the model can realistically input facial data into other videos (Sari, 2024a). b. Generative Adversarial Networks (GANs): Discriminators and generators are the building blocks of GAN. The discriminator learns to distinguish between real and fake data, and the generator generates new data closer to the accurate data. GAN builds a model that can develop material that looks or sounds like it was produced by someone else using actual video or audio as training data. This creates an environment that favors spreading false information and possible harassment, which can harm society (Sari, 2024b). The first use of deepfakes was in 2017, and it became more serious after the launch of the FakeApp app in 2018, which allows users to edit and swap faces between people with video output. Unfortunately, the negative impact of deepfakes is alarming. According to data from Home Security Heroes, 95,820 deepfake videos were distributed worldwide in 2023. Digitally altering sounds, images, and videos to portray a message that never happened in reality can be done with deepfake techniques (Patria, 2024). The development of digital technologies, especially in image and video manipulation, has seriously affected the integrity of information and the reputation of bodies and individuals. It is easy for these technologies to spread misleading information, influence public opinion, and even spark civil unrest. Unlike humans, technology has no boundaries against which to act and make decisions. Despite its extraordinary ability to process data or recognize complex patterns, AI cannot determine w (...truncated)