EU’s AI Code of Practice Reveals Industry Divide Over Regulatory Framework for General-Purpose AI Models

Title: The EU’s AI Code of Practice: A Divisive Guide to Regulating Artificial Intelligence

The European Union’s Artificial Intelligence (AI) Code of Practice has elicited contrasting responses from technology companies, with some entities expressing support for the framework, while others have deemed it as an instance of regulatory overreach.

The code functions as a pragmatic implementation guide for the EU AI Act, a groundbreaking legislation that seeks to regulate AI systems according to their potential risks, making it the world’s first comprehensive AI legislation.

The act introduces a risk-based framework that imposes more stringent regulations on AI systems perceived to pose greater threats to safety, fundamental rights, and society. However, the voluntary Code of Practice focuses specifically on general-purpose AI models, which are versatile systems like those developed by OpenAI, Google, and Meta, capable of performing multiple tasks rather than being designed for specific applications.

This emphasis on general-purpose AI models is significant given their power and potential impact on various sectors and societal aspects. The Code of Practice aims to provide guidance on the responsible development, deployment, and use of these systems, ensuring they are aligned with European values and promote trustworthy AI.