Microsoft has indicated it will likely become a signatory to the European Union´s newly established code of practice on artificial intelligence, marking a notable divergence from Meta´s position. Speaking to Reuters, Microsoft president Brad Smith noted the company needs to review the documents but expects to back the voluntary guidelines, designed to help enterprises adhere to the EU´s landmark artificial intelligence legislation enacted in June 2024. Smith highlighted a positive outlook on collaborating directly with the EU´s regulatory body for artificial intelligence.
The voluntary code, crafted by a panel of 13 independent experts, requires participating companies to publish detailed summaries on the data sets used to train general-purpose artificial intelligence models and establish policies ensuring compliance with EU copyright laws. This initiative forms part of the broader AI Act, whose scope covers giants such as Google (Alphabet), Meta, OpenAI, Anthropic, Mistral, and thousands of smaller firms. The code aims to provide legal clarity in an evolving regulatory landscape, yet its voluntary nature has drawn mixed responses across the industry.
Meta Platforms, however, has made clear it will not sign the code. Joel Kaplan, Meta’s chief global affairs officer, argued that the guidelines introduce significant legal ambiguity for model developers and impose requirements beyond those stipulated in the AI Act itself. In a blog post, Kaplan aligned Meta with 45 European companies who fear these measures may stifle the growth and experimentation required for frontier artificial intelligence model development in Europe. Despite Meta´s resistance, OpenAI and Mistral have already agreed to the code, illustrating growing divisions among major technology players as Europe seeks to chart a new course for artificial intelligence regulation and responsible deployment.
