Microsoft signals intent to join EU AI code as Meta declines
Microsoft is poised to become a signatory of the European Union´s recently introduced code of practice designed to guide companies in complying with the bloc´s new artificial intelligence regulations. The company´s president, Brad Smith, indicated in an interview with Reuters that Microsoft is inclined to endorse the voluntary framework, which aims to offer legal clarity for firms deploying general-purpose artificial intelligence models across Europe. The code, crafted by a panel of 13 independent experts, obliges participating companies to publish summaries of training data used for their artificial intelligence models and to implement policies that align with EU copyright law.
The code of practice ties directly to the EU´s Artificial Intelligence Act, which came into force in June 2024. The legislation targets major industry players including Alphabet, Meta, OpenAI, Anthropic, and Mistral, covering thousands of organizations operating within the bloc. As the regulatory environment evolves, the European Union´s artificial intelligence office has engaged directly with technology companies, seeking collaboration to ensure practical implementation while safeguarding innovation and compliance with legal standards. Brad Smith expressed Microsoft´s appreciation for this engagement and reaffirmed the company´s intention to act supportively pending a review of the documentation.
Meta Platforms, however, has reiterated its opposition to the code. Joel Kaplan, Meta´s chief global affairs officer, stated the company would not sign the code, accusing it of introducing legal ambiguities and imposing requirements that exceed the provisions of the Artificial Intelligence Act itself. He highlighted shared concerns with a coalition of 45 European firms, warning that these measures could stifle development and commercialization of advanced artificial intelligence models in Europe, hindering local technology businesses. In contrast, companies such as OpenAI and Mistral have already signed the code, signaling differing approaches to regulatory adaptation among leading artificial intelligence providers.