The European Union has released the final text of its groundbreaking Law on Artificial Intelligence, set to come into effect on August 1. This phased legislation addresses the ethical use of AI, transparency requirements, and copyright compliance, shaping the future of AI development.

Introduction


The European Union has taken a significant step in regulating artificial intelligence with the publication of the final text of its Law on Artificial Intelligence. Set to come into effect on August 1, this comprehensive legislation aims to ensure ethical and transparent use of AI technologies. The law will be implemented in phases, with some provisions taking effect immediately and others rolling out over the next two years.

A Phased Approach to AI Regulation

Immediate Bans on Misuse


Within six months of its enforcement, the EU will begin implementing bans on the illegal use of artificial intelligence. This includes prohibiting social rating systems similar to China’s social credit system, restricting the collection and compilation of facial recognition data, and banning real-time emotion recognition systems in schools and workplaces. These measures aim to protect citizens' privacy and prevent misuse of AI technologies.

Developing a Code of Practice


Nine months after the law comes into effect, the EU will start implementing a code of practice for AI developers. The European Commission's Office for Artificial Intelligence will collaborate with consulting companies to develop these guidelines. However, there are concerns that industry giants might influence these rules, potentially undermining the legislation's intent to regulate them.

Anti-Deepfake Measures and Transparency Requirements

New Rules for General-Purpose AI


One year after the law's enactment, developers of general-purpose AI models, such as Chat-GPT, will be required to meet new transparency standards. These standards mandate that AI systems be secure, explainable, and user-friendly. Additionally, the AI Act introduces rules for generative AI and manipulated media, requiring clear labeling of deepfakes and other AI-generated content. These measures aim to enhance user trust and prevent misinformation.

Ensuring Copyright Compliance

Balancing Innovation and Rights Protection


The new AI law also addresses copyright issues for companies training their AI models. Developers must comply with copyright laws unless their models are created solely for research purposes. The legislation states, “Rights holders may retain their rights over their works or other objects to prevent the extraction of text and data, unless this is done for scientific research purposes.” This provision aims to balance the protection of intellectual property with the need for scientific advancement.

Conclusion


The EU's Law on Artificial Intelligence marks a crucial step in regulating the rapidly evolving field of AI. By implementing a phased approach, the EU aims to address the ethical, transparency, and copyright issues associated with AI development. As these regulations come into effect, they will shape the future of AI, ensuring that its use benefits society while safeguarding individual rights and promoting innovation.