Meta’s Refusal to Sign EU’s AI Code of Practice: Implications and Insights
Meta, under the leadership of its chief global affairs officer Joel Kaplan, has officially declined to endorse the European Union’s recently introduced Code of Practice for its AI Act. With the regulations set to come into effect soon, this bold move raises questions about the future of AI development in Europe and beyond.
The Code of Practice Explained
Published early this month, the EU’s Code of Practice aims to steer companies towards compliance with the bloc’s comprehensive legislation on regulating artificial intelligence (AI). Designed as a voluntary framework, the code outlines crucial directives for implementing robust processes and systems while utilizing AI technologies.
One significant requirement is that companies must create and regularly update thorough documentation relating to their AI tools and services. Additionally, developers are prohibited from training their AI systems using pirated content, and they must honor requests from content owners not to include their works in training datasets. These measures aim to ensure ethical and legal compliance within the realms of AI development.
Meta’s Concerns and Criticism
In a recent LinkedIn post, Kaplan articulated Meta’s stance, labeling the EU’s regulatory framework as potentially obstructive. He stated, “Europe is heading down the wrong path on AI,” asserting that the Code introduces substantial legal uncertainties for model developers. According to him, these requirements exceed the intended scope of the AI Act, consequently hampering the innovative potential of AI technologies in Europe.
Kaplan described the EU’s implementation of this legislation as “overreach,” arguing that the Code could significantly inhibit the development and deployment of advanced AI models, as well as stifle European businesses aspiring to innovate in this burgeoning field.
More Read
Understanding the AI Act’s Provisions
The AI Act proposes a risk-based regulation approach for various AI applications. Certain use cases deemed “unacceptable risks” are outright banned, including forms of cognitive behavioral manipulation and social scoring. The Act also categorizes several AI applications as “high-risk.” This includes technologies around biometrics, facial recognition, and applications within sensitive domains like education and employment.
To ensure compliance, developers are required to register their AI systems, adhering to rigorous risk and quality management protocols. These measures aim to protect users and society at large, while establishing accountability for AI developers.
Industry Pushback and Future Outlook
Meta is not alone in its reluctance to embrace the EU’s regulatory measures. Major tech companies, including Google’s parent company Alphabet, Microsoft, and Mistral AI, have actively challenged these emerging regulations. These companies have even petitioned the European Commission for a delay in its implementation schedule, viewing it as a hindrance to innovation.
Despite this pushback, the European Commission remains steadfast in its commitment to the established timeline. With new guidelines recently released for AI model providers ahead of the rules going into effect on August 2, companies like OpenAI, Anthropic, Google, and Meta must face the reality of compliance. Any general-purpose AI models available prior to this date will need to adhere to the legislation by August 2, 2027.
Key Takeaways for AI Developers
Meta’s refusal to sign the EU’s Code of Practice signals a critical juncture in how leading AI developers navigate regulatory landscapes. As Europe reinforces its commitment to responsible AI development, understanding the implications of the AI Act becomes crucial for stakeholders in the industry. The call for clear and adaptive regulations that foster innovation rather than stifle it remains central to ongoing discussions within the tech community.
With rising tensions between regulatory bodies and tech giants, the unfolding landscape of AI governance will be one to watch closely in the coming years. As the world accelerates toward increasingly advanced AI technologies, how these regulations are implemented and received will reshape the framework for AI development for years to come.
Inspired by: Source
![Understanding How Learning Rate Decay Can Waste Valuable Data in Curriculum-Based LLM Pretraining: Insights from [2511.18903]](https://aimodelkit.com/wp-content/uploads/2026/04/1777269648_arxiv-logo-fb-150x150.png "Understanding How Learning Rate Decay Can Waste Valuable Data In Curriculum-Based Llm Pretraining: Insights From [2511.18903]"){kind=logo}
