Meta Snubs EU’s AI Code of Practice: Stirs Controversy Over Ethical Artificial Intelligence Standards

Meta has opted not to endorse the European Union’s Code of Practice for its AI Act, a decision made weeks prior to the implementation of the bloc’s regulations governing general-purpose artificial intelligence (AI) providers.

In a LinkedIn post, Joel Kaplan, Meta’s Chief Global Affairs Officer, expressed concern about Europe’s approach to AI regulation. He stated that Meta would not be signing the European Commission’s Code of Practice for General-Purpose AI (GPAI) models due to its introduction of legal uncertainties and measures extending beyond the scope of the AI Act.

The EU’s code of practice, a voluntary framework released earlier this month, is designed to assist companies in adhering to the bloc’s legislation for AI regulation. The code mandates companies to disclose and regularly update documentation regarding their AI tools and services; prohibits training AI on pirated content; and obliges compliance with content owners’ requests to exclude their works from datasets.

Labeling the EU’s legislative implementation as an “overreach,” Kaplan argued that the law would impede the advancement and deployment of cutting-edge AI models in Europe, thereby hampering European companies aiming to establish businesses based on these technologies.

The AI Act represents a risk-based regulation for AI applications, prohibiting certain uses associated with “unacceptable risk,” such as cognitive behavioral manipulation or social scoring. The rules also classify specific high-risk applications like biometrics and facial recognition in sectors including education and employment. Additionally, the act requires developers to register AI systems and adhere to risk and quality management obligations.

Global tech companies, including industry leaders such as Alphabet, Meta, Microsoft, and Mistral AI, have been vocal in their opposition to these rules, even urging the European Commission to delay implementation. However, the Commission has maintained its position, asserting that it will not alter its timeline.

On Friday, the EU also published guidelines for providers of AI models ahead of regulations set to take effect on August 2. These regulations will impact providers of “general-purpose AI models with systemic risk,” encompassing entities like OpenAI, Anthropic, Google, and Meta. Companies possessing such models before August 2 will be required to adhere to the legislation by August 2, 2027.