Meta Opts Not to Sign Up to EU AI Code of Practice as New AI Rules Loom

In a decision that reignites a simmering debate over Big Tech’s obligations in the era of artificial intelligence, Meta has now formally declined to sign Europe’s code of practice connected to the looming AI Act. The move comes just weeks before the European Union’s historic restrictions on general-purpose AI (GPAI) models are supposed to come into force.

Meta’s move sets it in opposition to European regulators and underscores the rising tension between global tech giants and governmental review in the field of AI.

A Pivotal Moment for AI Regulation in the EU

The EU’s AI Act—its first significant piece of regulation addressing artificial intelligence—is set to begin affecting general-purpose AI providers this summer.

• The code of conduct, while currently voluntary, serves as a middle ground between regulators and AI companies.
• It pushes organizations to demonstrate transparency, safety, and responsibility in the development and deployment of AI on a large scale.

Significant players in AI—including OpenAI, Google, Microsoft, and Anthropic—have signed the code, signaling their willingness to align with European norms.

Meta’s decision not to sign not only kicks it away from the pack but also calls into question its dedication to safe AI development in Europe.

Meta’s Justification: A Different Strategy

Meta stated that its refusal is due to:

“A preference for a more stable regulatory environment” rather than a voluntary, non-binding code.

• A Meta spokesperson emphasized that while the company endorses the goals of the AI Act, it believes compliance should be based on enacted legislation, not transitional deals.
• Meta is waiting until legally binding obligations of the AI Act take effect before making public commitments.

This method is not unprecedented—tech companies have previously delayed action in the face of regulatory uncertainty.

Critics warn that such delays could compromise the safety and transparency of AI tools in the interim.

EU’s Response: Disappointment and Determination

European regulators have expressed dissatisfaction with Meta’s stance.

Thierry Breton, EU’s Internal Market Commissioner and co-architect of the AI Act, remarked:

“This is a gradual process that is not a gesture, a communication only, but this is just the beginning.”

“It is a proof of confidence and a step toward making sure AI will be developed the way it fits to our European values: fairness, safety and trust,” she said.

• Officials suggest Meta’s refusal may impact how closely it will be monitored and assessed once the AI Act takes effect.
• Though voluntary, signing the code is viewed as a signal of good faith—a factor that may influence how Meta’s compliance is evaluated later.

A Complex Regulatory Landscape

Meta’s decision comes amid growing global complexity in AI regulation:

• The EU’s AI Act joins other frameworks like:
  • The Biden administration’s AI Executive Order in the U.S.
  • China’s centralized approach to AI governance.

Meta’s strategic positioning is especially important because of its involvement in open-source AI models like the LLaMA series.

Open-sourcing can democratize technology but also increases the risk of misuse—a core concern for EU regulators.

By not signing the code, Meta avoids immediate obligations such as:

• Providing risk assessments
• Sharing safety testing data
• Implementing publicly verifiable safeguards

This grants Meta more flexibility, but also draws greater scrutiny from regulators and the public.

Industry Implications: Trust and Transparency on the Line

Meta’s position has raised alarms among:

• Digital rights groups
• AI ethicists

These groups argue that Meta’s stance could erode public trust in its AI products. With generative AI increasingly influencing:

• Personalized content delivery
• Automated moderation
• Content creation like publications, eBooks, and articles

…transparency, accountability, and bias mitigation are more important than ever.

By not signing, Meta may appear less cooperative, which could damage its reputation in Europe’s privacy-conscious market.

Yet, some industry analysts view Meta’s move as pragmatic:

• Committing now could mean expensive compliance revisions later if the final laws differ from the current code.
• Thus, Meta’s approach may be one of regulatory caution, not resistance.

What Comes Next?

Even without signing the code, Meta will soon face mandatory legal requirements:

• The AI Office has been established by the European Commission to:
  • Monitor compliance
  • Conduct risk assessments
  • Impose penalties for non-compliance

Meta will be expected to:

• Demonstrate transparency in AI model development
• Address potential harms
• Maintain clear documentation of its AI systems

Avoiding early scrutiny may benefit Meta in the short term, but it won’t protect it from long-term accountability.

A Test Case for AI Regulation

Ultimately, Meta’s decision not to sign the EU’s AI code of practice is about more than corporate positioning—it is a turning point in the global effort to govern artificial intelligence.

As governments race to regulate, tech companies are being called to lead not just in innovation but in public trust-building.

Whether Meta’s bet is prudent or short-sighted remains to be seen. But one thing is certain:

The road to responsible AI is paved not only with laws and penalties, but also with voluntary commitments, ethical decisions, and a global expectation of transparency and care.

In this evolving landscape, every decision — even the decision not to sign — matters.