The European Union has pressed forward with the implementation of its new AI Act, a significant development impacting major U.S. tech firms such as Meta Platforms (META) and Alphabet-owned Google (GOOGL), alongside European group ASML (ASML). In a crucial step, the EU released a new code of practice aimed at guiding thousands of businesses within the 27-nation trading bloc in complying with its comprehensive artificial intelligence regulations. This code specifically addresses general purpose AI technology, including widely used chatbots like OpenAI’s ChatGPT.

The voluntary code of practice, which complements the broader EU AI Act, focuses on three primary areas: enhancing transparency requirements for providers of AI models seeking to integrate their technologies into various products; establishing robust copyright protections for AI-generated content; and ensuring the safety and security of the most advanced AI systems. The EU AI Act itself, approved last year and being phased into effect, imposes stringent regulations. It explicitly bans practices such as cognitive behavioral manipulation and social scoring, while also defining a set of “high-risk” uses for AI. These high-risk applications encompass sensitive domains like biometrics and facial recognition, as well as AI deployed in critical sectors such as education and employment. App developers creating systems categorized as high-risk will be mandated to register their systems and adhere to strict risk and quality management obligations to gain access to the EU market. Non-compliance with these regulations could lead to substantial penalties, including fines of up to 35 million euros ($41 million) or 7% of a company’s global revenue, whichever is higher.

While the full scope of the rules is set to come into force by mid-2026, certain AI applications, such as chatbots, are considered lower risk and will consequently face fewer compliance obligations. Despite the EU’s efforts to establish a regulatory framework for AI, the new regulations have drawn criticism from some corners of the industry. Major technology companies, including Meta, have voiced concerns, labeling the regulations as potentially “unworkable.” This sentiment is echoed by over 40 European companies, including aerospace giant Airbus (EADSF) and the French AI startup Mistral, which collectively urged the bloc in an open letter to postpone the full implementation of the regulations for two years. They argue that additional time is necessary to simplify what they describe as “unclear, overlapping and increasingly complex EU regulations” that could potentially undermine Europe’s competitiveness in the global AI landscape. However, the European Commission views the code of practice as a positive step. Henna Virkkunen, the commission’s executive vice president for tech sovereignty, security, and democracy, stated that the publication of the final version of the Code of Practice for general-purpose AI marks an important stride in ensuring that advanced AI models available in Europe are not only innovative but also safe and transparent.