Google Adopts EU AI Code as Meta Resists Over Innovation Concerns

Google has committed to signing the European Union’s voluntary AI code of practice, aligning with OpenAI and French startup Mistral in supporting the bloc’s regulatory framework despite reservations about its long-term impact on AI innovation [1]. Kent Walker, Alphabet’s president of global affairs and chief legal officer, emphasized the company’s conditional support, highlighting concerns that the code could slow Europe’s AI development and expose trade secrets through transparency requirements [2]. The code, designed to guide compliance with the EU’s AI Act set to take effect in August 2025, aims to harmonize practices among developers of advanced models like Gemini and Llama [1].

Meta, however, has refused to adopt the code, with chief legal officer Joel Kaplan criticizing it as introducing “legal uncertainties” and overreaching beyond the AI Act’s scope [3]. Kaplan argued the guidelines risk stifling innovation by imposing regulatory burdens on developers and limiting Europe’s ability to compete globally [2]. The company’s stance reflects broader resistance from tech firms and European industries, including Airbus and BNP Paribas, which have lobbied for a temporary pause on AI legislation citing overlapping and unclear rules [1].

The EU’s AI Act, enacted in March 2024, mandates risk assessments and transparency measures for general-purpose AI systems. While the code of practice is not legally binding, it serves as a voluntary blueprint for compliance. Companies like Google are expected to follow its guidelines ahead of the Act’s enforcement, signaling strategic alignment with Brussels’ regulatory priorities [1]. Microsoft has not yet signed the code but hinted at a likely decision, while the U.S. government has indirectly pressured the EU to soften its approach, framing its rules as “unjustified digital trade barriers” [1].

Legal experts acknowledge the challenges of implementing the code for rapidly evolving technologies. Nils Rauer of Pinsent Masons noted the guidance’s “generally well-prepared” nature but warned its generic framework could hinder practical implementation [4]. The delayed publication of the code—originally scheduled for May—has added uncertainty, with member states yet to formally endorse it. Meanwhile, the EU remains firm in its commitment to autonomous regulation, rejecting U.S. calls to revise its stance [1].

The divide between Google and Meta underscores the tension between regulatory caution and innovation. Google’s participation positions it as a key partner in shaping the EU’s AI governance, while Meta’s refusal highlights concerns over compliance costs and competitiveness. The Act’s phased implementation, with stricter rules for high-risk systems in 2026, further pressures companies to adapt. Legal teams across sectors, from pharmaceuticals to automotive, are now grappling with updating contracts and supply chains to meet the new obligations [4].

As the EU seeks to establish itself as a global leader in AI safety, the contrasting strategies of major tech firms reflect broader debates over balancing accountability with progress. Google’s proactive alignment with the code contrasts with Meta’s reluctance, illustrating the complex dynamics between regulators and industry stakeholders in defining the future of AI governance.

Sources:

[1] https://coinmarketcap.com/community/articles/6889e4db54b60d73c6013def/

[2] https://www.reuters.com/sustainability/boards-policy-regulation/google-sign-eus-ai-code-practice-despite-concerns-2025-07-30/

[3] https://www.capacitymedia.com/article-google-signs-eu-ai-code-of-practice

[4] https://www.theregister.com/2025/07/30/eu_ai_act/