Meta rejects EU AI Code of Practice, calling it a threat to growth
Meta’s global affairs chief Joel Kaplan warned that the guidelines could hinder innovation and the growth of AI businesses in Europe.
Meta Platforms has announced that it will not sign the European Union’s new Code of Practice for General-Purpose AI, arguing that the framework overreaches and may hinder innovation. In a statement posted on LinkedIn, Meta’s global affairs chief Joel Kaplan stated that the EU’s approach introduces legal uncertainties for model developers and imposes obligations that go beyond the requirements set out in the AI Act, which was adopted by European lawmakers in 2024. Kaplan expressed concern that the guidelines could negatively impact companies developing or building upon AI models in Europe.
The European Commission released the final version of the voluntary code last week. It is intended to help companies align with the AI Act by offering guidance on improving transparency and safety in the development and deployment of AI systems. The code is expected to come into effect in August 2025, ahead of the AI Act’s enforcement.
Meta’s decision comes amid broader industry debate. While OpenAI has agreed to sign the code, other major firms, including ASML and Airbus, have urged the European Commission to delay implementation by two years, citing concerns about the potential impact on innovation. Kaplan echoed these views, stating that the EU risks undermining the development of frontier AI models and the businesses that depend on them.
