Meta Rejects EU AI Code, Citing Legal Risks and Innovation Concerns

Introduction

On July 19, 2025, Meta publicly announced its decision not to sign the European Union’s newly introduced AI code of practice. This voluntary framework is designed to complement the EU's AI Act by providing detailed guidelines for companies developing and deploying AI technologies in Europe. Meta’s refusal to endorse the code is a significant development in the growing debate over AI regulation, reflecting broader industry concerns about legal ambiguity and innovation constraints.

Key Details

The EU’s AI code of practice, published on July 10, 2025, aims to operationalize provisions of the AI Act by setting standards on data usage, transparency, and compliance.
Meta’s Chief Global Affairs Officer, Joel Kaplan, described the code as an "overreach" that creates legal uncertainties beyond the scope of the AI Act.
The code requires companies to avoid using pirated materials in AI training and respect artists’ and writers’ requests to exclude their work from datasets.
Signing the code is voluntary but offers legal protections against accusations of breaching the AI Act.
Non-signatories like Meta may face increased regulatory scrutiny and must demonstrate compliance through other means.
Violations of the AI Act can result in fines up to 7% of annual global turnover for general infringements, and 3% for advanced AI models.

Background

The EU’s AI Act represents one of the world’s first comprehensive legal attempts to regulate AI technologies, aiming to balance innovation with fundamental rights protection. The AI code of practice serves as a supplementary voluntary mechanism to help companies navigate regulatory requirements. Meta’s position aligns with ongoing skepticism from major US tech firms, which argue that Europe’s regulatory approach is unpredictable, overly restrictive, and detrimental to technological progress.

Meta’s resistance is not new. Earlier in 2025, company executives publicly criticized the AI Act for hampering innovation and delaying product launches, warning that European consumers may ultimately suffer from limited access to advanced AI solutions. The company’s stance also reflects broader transatlantic tensions on digital regulation, with the US administration under then-President Trump openly pressuring the EU to abandon or weaken its AI regulatory framework.

Impact Analysis

Meta’s refusal to sign the EU AI code of practice signals a significant challenge for European regulators. As one of the largest AI developers globally, Meta’s non-participation could hinder the uniform adoption of ethical AI standards. The company’s critique of legal uncertainties highlights a risk that regulatory frameworks may inadvertently stifle innovation by creating compliance ambiguities.

Moreover, Meta’s public rejection could influence other AI firms to adopt a similar stance, potentially fragmenting the regulatory landscape within Europe. This fragmentation complicates enforcement and may reduce the effectiveness of the AI Act’s protections for citizens and consumers.

Joel Kaplan stated, "Europe is heading down the wrong path on AI. We have carefully reviewed the European Commission’s Code of Practice for general-purpose AI models and Meta won’t be signing it. This Code introduces a number of legal uncertainties for model developers, as well as measures which go far beyond the scope of the AI Act."

Broader Context

The EU’s AI regulatory initiative emerges amid a global race to develop trustworthy and safe AI. Europe positions itself as a leader in AI ethics, emphasizing transparency, accountability, and safeguarding individual rights. However, this approach contrasts with the more innovation-driven, lighter-touch regulatory strategies favored by many US companies and policymakers.

Meta’s opposition reflects a broader debate about balancing regulation and innovation. Critics argue that stringent rules can slow AI development, reduce competitiveness, and discourage investment. Supporters contend that robust oversight is essential to prevent misuse, bias, and harmful societal impacts.

Adding complexity, the political climate in the US under the Trump administration has exhibited skepticism toward AI regulation, viewing it as burdensome to business. This geopolitical dynamic underscores the challenges of creating harmonized international standards for AI governance.

Future Outlook

The EU is likely to continue enforcing its AI Act rigorously, with or without Meta’s participation in the voluntary code. Companies refusing to sign the code will face higher scrutiny and must demonstrate alternative compliance methods. This may push Meta to increase its engagement on specific regulatory dialogues or develop independent compliance frameworks.

For the AI industry, Meta’s stance may trigger a deeper dialogue on how to balance innovation with ethical and legal safeguards. Other companies may weigh the benefits of signing the code for legal protections against concerns about regulatory overreach.

Internationally, the divergence in approaches between the EU and US may complicate cross-border AI development and deployment, necessitating ongoing negotiations to align standards and reduce friction.

Conclusion

Meta’s decision not to sign the EU’s AI code of practice underscores the tensions between leading AI developers and regulatory bodies striving to control the technology’s risks. While the code aims to foster ethical AI development within Europe, Meta’s rejection highlights challenges in creating clear, workable frameworks that satisfy both innovation imperatives and regulatory safeguards. The evolving dialogue between regulators and AI firms will be critical in shaping the future landscape of artificial intelligence governance globally.