Google has agreed to sign the European Union’s voluntary AI Code of Practice, positioning itself alongside other major AI players like Microsoft, OpenAI, Anthropic, and Mistral. Meanwhile, Meta—the only major U.S.-based AI company to decline—remains on the sidelines, raising sharp concerns over the code’s legal and operational implications.
Google Says Yes, But with Caveats
Despite voicing concern that the EU’s regulatory approach could hinder AI innovation, Google confirmed it will adopt the guidelines. The company framed the decision as a step toward maintaining European access to “secure, first-rate AI tools.”
Behind the scenes, however, Google remains wary. The tech giant has warned the EU code, along with the broader AI Act, risks entangling AI development in regulatory delays, exposing trade secrets, and causing legal ambiguity around copyright compliance. In particular, it fears Europe’s competitiveness could suffer if the balance between safety and innovation tips too far toward bureaucracy.
Meta Draws a Line
Meta’s leadership sees things differently. Joel Kaplan, Meta’s global policy chief, publicly rejected the Code, calling it an overreach that adds confusion rather than clarity. Kaplan argues the EU’s voluntary framework “stretches beyond” the AI Act, imposes legal uncertainty, and may “stunt innovation” across the continent.
With that, Meta becomes the lone holdout among top AI firms operating in Europe.
What’s in the Code?
The EU AI Code of Practice is not legally binding, but it’s designed to prepare companies for the enforcement of the AI Act, parts of which begin to take effect this month. It covers:
- Transparency: Disclosing model capabilities, limitations, and risks
- Copyright compliance: Respecting IP rights in model training and deployment
- Safety and security: Preventing harmful misuse or unintended consequences
By signing, companies gain a more predictable regulatory path and possibly smoother interactions with EU oversight bodies.
A Fault Line in the Industry
The split between Google and Meta reflects deeper tensions in how U.S. companies view European tech regulation. Google is choosing engagement—hoping to shape the rules from within. Meta is betting that resistance will buy it more flexibility, or that the EU will blink first in a high-stakes negotiation over AI’s future.
With the AI Act set to become fully enforceable by 2026, the choices these companies make now will shape how—and where—they can deploy their most powerful AI tools.
The Road Ahead
The EU says the Code is just a starting point. Lawmakers and regulators are watching closely to see how the voluntary framework plays out—and whether firms like Meta will face added scrutiny for opting out.
For now, Europe has drawn a line. Most companies are stepping toward it. Meta is standing apart.