by : Sonu Kumar, 4th year(8th Semester) BBA LL.B., Lovely Professional University, Phagwara, Punjab.
Abstract
The European Union (EU) has recently passed a ground-breaking EU AI Act, the first ever all-embracing law that talks about Artificial Intelligence (AI), its risks and ethical considerations. This article will highlight some of the main points about the Act, which include its risk-based approach to regulation, categorization of AI systems as well as other measures concerning deep fakes and transparency. By discussing how the EU is leading in this area, this article lays that it could serve as a template for global AI governance supporting responsible development anchored on human rights protection and innovation promotion.
Keywords
Artificial Intelligence (AI), EU AI Act, AI regulation, risk-based approach, deep fakes, transparency, ethics, global AI governance
Introduction
Artificial intelligence (AI) is the automated robot which stores vast amount of data and can analyze and work on that data in lightning speed. The rapid advancement of technology has brought about an era where smart machines are revolutionizing aspects of our daily lives. From vehicles, to automated decision-making processes these innovations offer potential. Due to its rapidness in analysing data, the growth of AI worldwide is significant. However, this progress calls for a review of the frameworks that regulate their creation and use. This opening section delves into the intricacies of establishing guidelines for technologies. We will examine the shortcomings of laws in addressing the challenges posed by these advanced tools. The chapter will then outline principles for promoting innovation taking inspiration from the pioneering work of the European Union’s AI Act.
Given the unprecedented rate at which AI technology is advancing, legislatures, legal practitioners and decision makers face a difficult task of coming up with a strong legal framework that can successfully deal with complex ethical, sociological and financial issues associated with AI.
While artificial intelligence (AI) systems become more autonomous and capable of making decisions on their own, there are worries about accountability for such decisions or outcomes. Additionally, strict scrutiny from the law should be upheld because ethics are involved in the development of AI such as justice, transparency and accountability. Both sides have initiatives working toward setting norms and standards for AI governance; these include the General Data Protection Regulation (GDPR) in Europe through to other organizations like IEEE or even Partnership on AI having principles or guidelines surrounding ethics.However, as new sectors imbibe AI technologies while it continually evolves, legal frameworks must remain agile and adaptive in recognition of emerging challenges effectively mitigating against potential risks.
Historic day (March 13, 2024) for the European Union when it passed the EU Law on Artificial Intelligence.[1] The European Union Law on Artificial Intelligence acts as a comprehensive framework for dealing with AI risks. It is the first law maker with respect to AI in the world. Europe says it will enable innovation while protecting fundamental rights and at the same time it will focus on technology advancement under human control.
Governance of New law with respect to AI and its dealing with deep fakes.
The law regulating AI in the European Union is called the EU AI Act, a comprehensive law on artificial intelligence essentially the AI Control Framework and is passed. The AI law is a first for the European Union and a first for the world, it was approved by the European Parliament and now only needs the formal approval of the EU states. ACT has a simple principle, i.e. it wants to make technology human-centric, so the idea is to regulate AI based on its ability to harm society, the higher the risk, the stricter the rules. First, let's look at how ACT defines artificial intelligence. The basic definition of artificial intelligence is computers that do tasks and solve problems like humans, but this act takes a more detailed approach, defining it as a machine system designed to operate with varying levels of autonomy, which is a lot of technical jargon, but basically it refers to the same things that also include chatbots like Chat GPT and Gemini.
Bills effect on Artificial Intelligence risks
ACT ranks AI systems according to how they can be low-risk, medium-risk, or high-risk. High risk systems include those deployed in banking, schools or critical infrastructure and must be accurate. One must overlook them and their use must be monitored if they directly affect citizens. They have the right to question these systems, then there are systems that are prohibited, i.e. artificial intelligence systems that can harm, can affect people's lives, for example - a social scoring system, basically systems that classify people based on social behavior. or personality. China has one, but it's a strict no for the European Union, which brings us to the tax-exempt systems.[2] Not all AI tools fall under this law, for example, AI tools intended for military defense or national security are exempted, as well as tools intended for science and research. Facial recognition tools are allowed, but only for law enforcement.[3]
Important Concern of Deep Fakes
ACT aims to resolve that if content is artificially created or altered, it must be flagged. People, companies, authorities must mark it. Then there's generative AI, a type of AI that can create content, text images, audio, and basically anything. In the case of Gramophone Company of India Ltd. v. Super Cassettes[4] Industries Ltd. (2011) The Delhi High Court held that AI-generated music produced by a computer program lacked human creativity and was therefore ineligible for copyright protection. GPT chat types must meet specific requirements. Must cooperate with copyright law; must disclose the data they use to train these systems. Everything must be transparent.
Penalties for Violation of law
If the law is violated it has fines and they range from €7.5 million to €35 million. There are fines even for providing incorrect information to regulatory authorities, violating provisions of the law, and developing or deploying prohibited tools. It is going to affect the tech companies on a big scale and on this, tech companies response was mixed, some tech companies showed a very positive response and welcomed it while some had different opinions.[5] The Act will come into force around May, and proper implementation of this law may take time at least 2 years giving companies more time to adapt and adjust.[6] The EU set a trend for the AI law, but the US is also mandating AI developers to share data with the government. China has also introduced a patchwork of AI laws as we recognize the dangers of AI.
Conclusion
In conclusion, the European Union AI Act is a landmark for both EU and global society. It sets the stage for the responsible development of AI and prioritizes people’s well-being over technical advancement. However, other countries have already begun to regulate AI, but the EU’s expansive system offers a great example on how to balance between innovation and ethical concerns like deep fakes. In reviewing employment and enforcement of this legislation in years ahead are crucial. Achievement with respect to EU AI Act could see such laws being enacted globally ensuring that everyone benefits from artificial intelligence technologies guaranteeing that AI technologies are used for everyone's advantage.
[1] EU Parliament: https://www.europarl.europa.eu/portal/en
[2] "The European Union's Artificial Intelligence Act, explained" - World Economic Forum: https://www.weforum.org/agenda/2023/12/europe-landmark-ai-regulation-deal/
[4] Gramophone Company of India Ltd. v. Super Cassettes Industries Ltd. (2011)
[5] Partnership on AI - Partnership on AI website: https://partnershiponai.org/
[6] "(PDF) The EU AI Act: A pioneering effort to regulate frontier AI?" - ResearchGate: https://www.researchgate.net/publication/377180354_The_EU_AI_Act_A_pioneering_effort_to_regulate_frontier_AI